Pygor Tutorial

Welcome to the pygor3 Tutorial.

Pygor3 is an open source project and Python package that allows to analyze infer, evaluate and generate V(D)J sequences, by using IGoR’s.

Pygor3 could help you to get simple calculations and visualizations of the statistics in VDJ recombination

IgorModel

An IGoR model’s encapsulates the Bayesian network probabilistic parameters of a V(D)J recombination process. IGoR is shipped with a set of default models.

As an example lets load the recombination model for a human \(\beta\) T-cell receptor

import pygor3 as p3
mdl_hb = p3.get_default_IgorModel("human", "tcr_beta")

Reading Parms filename from:  /home/olivares/.local/share/igor/models/human/tcr_beta/models/model_parms.txt
Reading Marginals filename from:  /home/olivares/.local/share/igor/models/human/tcr_beta/models/model_marginals.txt

mdl_hb['d_3_del']

<xarray.DataArray (d_gene: 3, d_5_del: 21, d_3_del: 21)>
array([[[0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 3.19224e-01,
         2.89631e-01, 2.11165e-01],
        [0.00000e+00, 0.00000e+00, 6.86291e-08, ..., 1.38170e-01,
         3.02534e-01, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 1.09220e-03, ..., 4.41026e-02,
         0.00000e+00, 0.00000e+00],
        ...,
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00]],

       [[0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 2.11468e-03,
         5.71094e-03, 7.95666e-02],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 1.50811e-02,
         8.09057e-02, 7.76708e-01],
        [0.00000e+00, 0.00000e+00, 3.94418e-06, ..., 2.35577e-03,
         5.88810e-02, 7.62736e-02],
...
        [0.00000e+00, 0.00000e+00, 1.25405e-01, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00]],

       [[0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 1.51398e-04,
         1.89857e-02, 6.63111e-01],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 1.29730e-02,
         2.58015e-02, 9.39715e-01],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 1.20776e-02,
         1.25424e-01, 1.52907e-01],
        ...,
        [0.00000e+00, 0.00000e+00, 1.62236e-01, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00],
        [0.00000e+00, 0.00000e+00, 0.00000e+00, ..., 0.00000e+00,
         0.00000e+00, 0.00000e+00]]])
Coordinates:
  * d_gene        (d_gene) int64 0 1 2
    lbl__d_gene   (d_gene) object ' TRBD1*01' ' TRBD2*01' ' TRBD2*02'
    seq__d_gene   (d_gene) object 'GGGACAGGGGGC' ... 'GGGACTAGCGGGAGGG'
  * d_5_del       (d_5_del) int64 0 1 2 3 4 5 6 7 8 ... 13 14 15 16 17 18 19 20
    lbl__d_5_del  (d_5_del) int64 -4 -3 -2 -1 0 1 2 3 ... 9 10 11 12 13 14 15 16
  * d_3_del       (d_3_del) int64 0 1 2 3 4 5 6 7 8 ... 13 14 15 16 17 18 19 20
    lbl__d_3_del  (d_3_del) int64 -4 -3 -2 -1 0 1 2 3 ... 9 10 11 12 13 14 15 16
Attributes:
    nickname:    d_3_del
    event_type:  Deletion
    seq_type:    D_gene
    seq_side:    Three_prime
    priority:    5
    parents:     ['d_gene', 'd_5_del']
    childs:      []

Bayesian Network

To visualize the composition of the Bayesina network we can use

mdl_hb.plot_Bayes_network()

<AxesSubplot:>

Similarly we can do the same for a human \(\alpha\) T-cell receptor

mdl_ha = p3.get_default_IgorModel("human", "tcr_alpha")
mdl_ha.plot_Bayes_network()

Reading Parms filename from:  /home/olivares/.local/share/igor/models/human/tcr_alpha/models/model_parms.txt
Reading Marginals filename from:  /home/olivares/.local/share/igor/models/human/tcr_alpha/models/model_marginals.txt

<AxesSubplot:>

Notice that for a \(\beta\) T-cell receptor we get a VDJ model and for the \(\alpha\) only a VJ.

An IgorModel internally has two structure based in IGoR’s model files: - model_parms.txt that contains the information about the events name, nickname, priority, and the dependencies between each other. - model_marginals.txt which contains the conditional probabilities for each event. - CDR3 anchors files.

IgorModel_Parms

print(mdl_hb)

.xdata['v_3_del', 'j_choice', 'd_5_del', 'vd_dinucl', 'dj_dinucl', 'd_gene', 'v_choice', 'd_3_del', 'j_5_del', 'dj_ins', 'vd_ins']

mdl_hb.parms, mdl_hb.marginals

(<pygor3.IgorIO.IgorModel_Parms at 0x7fac64302950>,
 <pygor3.IgorIO.IgorModel_Marginals at 0x7fac6459d710>)

To get a pandas dataframe of the realizations of an event just use the nickname of the event.

mdl_hb.parms['d_gene']

	value	name
id
0	GGGACAGGGGGC	TRBD1*01
1	GGGACTAGCGGGGGGG	TRBD2*01
2	GGGACTAGCGGGAGGG	TRBD2*02

event = mdl_hb.parms.get_Event('d_gene')
event.to_dict()

{'event_type': 'GeneChoice',
 'seq_type': 'D_gene',
 'seq_side': 'Undefined_side',
 'priority': 6,
 'realizations': [{'id': 0, 'value': 'GGGACAGGGGGC', 'name': ' TRBD1*01'},
  {'id': 1, 'value': 'GGGACTAGCGGGGGGG', 'name': ' TRBD2*01'},
  {'id': 2, 'value': 'GGGACTAGCGGGAGGG', 'name': ' TRBD2*02'}],
 'name': 'GeneChoice_D_gene_Undefined_side_prio6_size3',
 'nickname': 'd_gene'}

All parameters of the events have 3 columns id, value and name. For the GeneChoice events like ‘d_gene’(above) the value is the sequence and the name the description of the sequence and for Deletion events like ‘d_3_del’ the value is an integer of the posible deletions.

mdl_hb.parms['d_3_del']

	value	name
id
0	-4
1	-3
2	-2
3	-1
4	0
5	1
6	2
7	3
8	4
9	5
10	6
11	7
12	8
13	9
14	10
15	11
16	12
17	13
18	14
19	15
20	16

Notice that for Deletion event negative values mean palidromic insertions

Aditionally the anchors dataframe are stores in parms.df_V_anchors for CDR3 V anchors and parms.df_J_anchors for J with gene (gene name) as index in

mdl_hb.parms.df_J_anchors

	anchor_index
gene
K02545|TRBJ1-1*01|Homo sapiens|F|J-REGION|749..796|48 nt|3| | | | |48+0=48| | |	17
K02545|TRBJ1-2*01|Homo sapiens|F|J-REGION|886..933|48 nt|3| | | | |48+0=48| | |	17
M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499..1548|50 nt|2| | | | |50+0=50| | |	19
M14158|TRBJ1-4*01|Homo sapiens|F|J-REGION|2095..2145|51 nt|3| | | | |51+0=51| | |	20
M14158|TRBJ1-5*01|Homo sapiens|F|J-REGION|2368..2417|50 nt|2| | | | |50+0=50| | |	19
M14158|TRBJ1-6*01|Homo sapiens|F|J-REGION|2859..2911|53 nt|2| | | | |53+0=53| | |	22
L36092|TRBJ1-6*02|Homo sapiens|F|J-REGION|643043..643095|53 nt|2| | | | |53+0=53| | |	22
X02987|TRBJ2-1*01|Homo sapiens|F|J-REGION|800..849|50 nt|2| | | | |50+0=50| | |	19
X02987|TRBJ2-2*01|Homo sapiens|F|J-REGION|995..1045|51 nt|3| | | | |51+0=51| | |	20
X02987|TRBJ2-3*01|Homo sapiens|F|J-REGION|1282..1330|49 nt|1| | | | |49+0=49| | |	18
X02987|TRBJ2-4*01|Homo sapiens|F|J-REGION|1432..1481|50 nt|2| | | | |50+0=50| | |	19
X02987|TRBJ2-5*01|Homo sapiens|F|J-REGION|1553..1600|48 nt|3| | | | |48+0=48| | |	17
X02987|TRBJ2-6*01|Homo sapiens|F|J-REGION|1673..1725|53 nt|2| | | | |53+0=53| | |	22
M14159|TRBJ2-7*01|Homo sapiens|F|J-REGION|2316..2362|47 nt|2| | | | |47+0=47| | |	16
X02987|TRBJ2-7*02|Homo sapiens|ORF|J-REGION|1890..1936|47 nt|2| | | | |47+0=47| | |	16

IgorModel_Marginals

IgorModel_Marginals contains the conditional probabilities for each event.

ATTENTION

IGoR’s model_marginals.txt stores conditional probabilities of the defined events, not marginals probabilities.

The conditional probability as a numpy array can be access with

mdl_hb.marginals['j_choice']

array([[1.28586e-01, 1.04003e-01, 4.10916e-02, ..., 0.00000e+00,
        5.36364e-02, 3.73186e-02],
       [1.66126e-01, 7.89615e-02, 1.13701e-02, ..., 2.46074e-02,
        6.49049e-02, 7.46140e-02],
       [1.66156e-01, 7.90103e-02, 1.13744e-02, ..., 2.45354e-02,
        6.49334e-02, 7.46554e-02],
       ...,
       [1.24050e-01, 4.76394e-02, 0.00000e+00, ..., 3.52196e-02,
        5.51254e-02, 6.01472e-02],
       [3.64941e-06, 2.04721e-09, 0.00000e+00, ..., 3.41728e-01,
        6.84855e-02, 8.22442e-02],
       [1.06077e-01, 1.13773e-01, 4.03129e-02, ..., 2.20679e-02,
        7.16947e-02, 5.81105e-02]])

DataArrays with parms and marginals

The mdl_hb object encapsulates the information about IGoR’s model, like the Bayes network and the corresponding conditional probabilities for each event and the information about the parms and marginals can be access directly from the xarray in IgorModel without using IgorModel_Parms and IgorModel_Marginals like

Conditional probabilities

To get the conditional probabilities associated with this model we can the DataArray as

IgorModel[‘event_nickname’]

For instance from this Bayesian network, we can see that for this model the choice of V (‘v_choice’) or the number of insertions between the V and D segments are independent of the rest of events. Hence,

\(P(\text{v_choice})\) = mdl_hb[‘v_choice’]

However, for events like ‘d_gene’ or ‘j_5_del’ there are some conditional dependencies, therefore the notation in pygor

\(P(\text{d_gene}| \text{v_choice}, \text{j_choice})\) = mdl_hb[‘d_gene’]

The get the dependencies information there is variable parents as attribute

mdl_hb['v_choice']

<xarray.DataArray (v_choice: 89)>
array([4.88741e-03, 9.32369e-03, 9.32259e-03, 1.30320e-02, 3.43430e-04,
       8.50694e-03, 7.71250e-03, 6.12276e-04, 5.06104e-03, 4.59289e-05,
       4.48245e-03, 8.16181e-03, 7.00053e-04, 7.77164e-03, 1.16174e-02,
       1.16158e-02, 1.13872e-02, 1.06555e-02, 2.36870e-03, 2.28368e-02,
       1.56715e-04, 4.58447e-03, 0.00000e+00, 4.88782e-05, 0.00000e+00,
       1.54500e-02, 2.74050e-02, 5.18979e-03, 4.80289e-03, 1.62765e-01,
       6.61229e-02, 2.07174e-02, 7.36226e-04, 2.47846e-37, 1.07472e-02,
       2.03820e-02, 9.53245e-03, 9.53006e-03, 2.35607e-03, 2.39208e-03,
       2.39208e-03, 2.45219e-03, 1.66904e-02, 2.98807e-03, 2.98804e-03,
       1.59156e-02, 1.25468e-02, 9.06711e-03, 9.06936e-03, 9.76534e-02,
       9.57746e-03, 9.17011e-03, 1.14472e-02, 1.14458e-02, 9.77554e-03,
       1.89924e-02, 3.97949e-04, 1.70242e-03, 8.91336e-03, 6.73926e-03,
       6.69167e-03, 2.94257e-02, 1.38534e-02, 4.19227e-03, 4.19275e-03,
       2.48630e-03, 4.83399e-04, 1.29795e-07, 6.49975e-05, 1.94736e-02,
       1.94728e-02, 0.00000e+00, 3.91595e-03, 3.96230e-02, 0.00000e+00,
       0.00000e+00, 6.73805e-05, 2.90073e-02, 9.28490e-03, 5.32118e-03,
       8.18432e-03, 1.07035e-03, 3.64764e-04, 3.26790e-03, 3.26801e-03,
       3.26801e-03, 3.26809e-03, 8.23600e-04, 1.56399e-02])
Coordinates:
  * v_choice       (v_choice) int64 0 1 2 3 4 5 6 7 ... 81 82 83 84 85 86 87 88
    lbl__v_choice  (v_choice) object 'U66059|TRBV1*01|Homo sapiens|P|V-REGION...
    seq__v_choice  (v_choice) object 'GATACTGGAATTACCCAGACACCAAAATACCTGGTCACA...
Attributes:
    nickname:    v_choice
    event_type:  GeneChoice
    seq_type:    V_gene
    seq_side:    Undefined_side
    priority:    7
    parents:     []
    childs:      ['v_3_del', 'd_gene', 'j_choice']

You can plot directly from the xarray

mdl_hb['v_choice'].plot()

[<matplotlib.lines.Line2D at 0x7fac3b5ef210>]

Or use the a model method to get the plot with the corresponding labels.

mdl_hb.plot_Event('v_choice')

(<Figure size 1296x1080 with 1 Axes>,
 <AxesSubplot:title={'center':'$P($v_choice$)$'}>)

mdl_hb.plot_Event('d_gene')

(<Figure size 720x1440 with 6 Axes>,
 array([<AxesSubplot:title={'center':'$P($d_gene$ = $  TRBD1*01 $|$v_choice,j_choice$)$'}, xlabel='j_choice', ylabel='v_choice'>,
        <AxesSubplot:title={'center':'$P($d_gene$ = $  TRBD2*01 $|$v_choice,j_choice$)$'}, xlabel='j_choice', ylabel='v_choice'>,
        <AxesSubplot:title={'center':'$P($d_gene$ = $  TRBD2*02 $|$v_choice,j_choice$)$'}, xlabel='j_choice', ylabel='v_choice'>],
       dtype=object))

Marginal Probabilities

With IGoR provide us the conditional probabilities of the events defined in the Bayesian network.

So we can calculate marginal probabilities, i.e.

\(P(\text{j_choice}) = \sum_{\text{v_choice}}P(\text{j_choice}, \text{v_choice})\)

and using the Bayes theorem

$ P(\text{j_choice}, \text{v_choice}) = P(\text{j_choice} | \text{v_choice}) \times `P(:nbsphinx-math:text{v_choice}`)$

we get,

\(P(\text{j_choice}) = \sum_{\text{v_choice}} P(\text{j_choice} | \text{v_choice}) \times P(\text{v_choice})\)

import xarray as xr
P_marginal_j_choice = xr.dot(mdl_hb['j_choice'], mdl_hb['v_choice'])
P_marginal_j_choice, P_marginal_j_choice.plot()

(<xarray.DataArray (j_choice: 15)>
 array([0.11962186, 0.10538077, 0.02016188, 0.05684014, 0.09945672,
        0.03726745, 0.0322361 , 0.12542452, 0.05034454, 0.09853381,
        0.01948227, 0.07500318, 0.0190534 , 0.0709002 , 0.07029376])
 Coordinates:
   * j_choice       (j_choice) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
     lbl__j_choice  (j_choice) object 'K02545|TRBJ1-1*01|Homo sapiens|F|J-REGI...
     seq__j_choice  (j_choice) object 'TGAACACTGAAGCTTTCTTTGGACAAGGCACCAGACTCA...,
 [<matplotlib.lines.Line2D at 0x7fac1dacf7d0>])

But when a model is loaded with pygor3 the marginals are calculated automatically using a variable elimination process and are store in the Pmarginal variable

IgorModel.Pmarginal[‘event_nickname’]

For this case

$ P(\text{'j_choice'}) = $ mdl_hb.Pmarginal[‘j_choice’]

mdl_hb.plot_Event_Marginal('d_gene')
mdl_hb.Pmarginal['d_gene']

<xarray.DataArray (d_gene: 3)>
array([0.56140871, 0.2361217 , 0.20247255])
Coordinates:
  * d_gene       (d_gene) int64 0 1 2
    lbl__d_gene  (d_gene) object ' TRBD1*01' ' TRBD2*01' ' TRBD2*02'
    seq__d_gene  (d_gene) object 'GGGACAGGGGGC' ... 'GGGACTAGCGGGAGGG'

mdl_hb.plot_Event_Marginal('d_3_del')

<AxesSubplot:xlabel='d 3 del', ylabel='P'>

If you modifiy the conditional probabilities, to re-calculate the marginals, use the method generate_Pmarginals()

mdl_hb.generate_Pmarginals()

Joint Probabilities

Pygor3 also have a method to calculate the joint probabilities of events

IgorModel.get_P_joint([‘event_nickname_1’, ‘event_nickname_2’, …])

WARNING

Be carefull with this function a the computer memory consumption could increase if more than 2 events are requested.

$P(\text{'v_choice'}, \text{'j_choice'}) = $mdl_hb.get_P_joint([‘v_choice’, ‘j_choice’])

P_V_J = mdl_hb.get_P_joint(['v_choice', 'j_choice'])
P_V_J

<xarray.DataArray (v_choice: 89, j_choice: 15)>
array([[6.28453463e-04, 5.08306114e-04, 2.00831906e-04, ...,
        0.00000000e+00, 2.62143710e-04, 1.82391677e-04],
       [1.54890932e-03, 7.36213918e-04, 1.06011576e-04, ...,
        2.29432410e-04, 6.05154843e-04, 6.95679479e-04],
       [1.54900602e-03, 7.36581450e-04, 1.06039046e-04, ...,
        2.28733827e-04, 6.05348578e-04, 6.95983268e-04],
       ...,
       [4.05407259e-04, 1.55690174e-04, 0.00000000e+00, ...,
        1.15101267e-04, 1.80155267e-04, 1.96566942e-04],
       [3.00565952e-09, 1.68608515e-12, 0.00000000e+00, ...,
        2.81447867e-04, 5.64048061e-05, 6.77364474e-05],
       [1.65903669e-03, 1.77940202e-03, 6.30491475e-04, ...,
        3.45140763e-04, 1.12130116e-03, 9.08844648e-04]])
Coordinates:
  * v_choice       (v_choice) int64 0 1 2 3 4 5 6 7 ... 81 82 83 84 85 86 87 88
    seq__v_choice  (v_choice) object 'GATACTGGAATTACCCAGACACCAAAATACCTGGTCACA...
  * j_choice       (j_choice) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
    seq__j_choice  (j_choice) object 'TGAACACTGAAGCTTTCTTTGGACAAGGCACCAGACTCA...

P_V_J.plot()

<matplotlib.collections.QuadMesh at 0x7fac1c1fc8d0>

mdl_hb['j_choice'].plot()

<matplotlib.collections.QuadMesh at 0x7fac1c0e7310>

Editing a model

A model can be edited manually if necessary. In cases like a long gene description or to add new anchors to the existing genomic templates

mdl_hb.genomic_dataframe_dict

{'V':                                                  name  \
 id
 0   U66059|TRBV1*01|Homo sapiens|P|V-REGION|91723....
 1   U66059|TRBV10-1*01|Homo sapiens|F|V-REGION|214...
 2   AF009660|TRBV10-1*02|Homo sapiens|F|V-REGION|5...
 3   AF009660|TRBV10-1*04|Homo sapiens|F|V-REGION|5...
 4   U66059|TRBV10-2*01|Homo sapiens|F|V-REGION|239...
 ..                                                ...
 84  M14261|TRBV7-9*04|Homo sapiens|(F)|V-REGION|58...
 85  M27385|TRBV7-9*05|Homo sapiens|(F)|V-REGION|34...
 86  X74844|TRBV7-9*06|Homo sapiens|(F)|V-REGION|10...
 87  L14854|TRBV7-9*07|Homo sapiens|(F)|V-REGION|1....
 88  U66059|TRBV9*01|Homo sapiens|F|V-REGION|206836...

                                                 value  anchor_index
 id
 0   GATACTGGAATTACCCAGACACCAAAATACCTGGTCACAGCAATGG...           NaN
 1   GATGCTGAAATCACCCAGAGCCCAAGACACAAGATCACAGAGACAG...         270.0
 2   GATGCTGAAATCACCCAGAGCCCAAGACACAAGATCACAGAGACAG...         270.0
 3   GATGCTGAAATCACCCAGAGCCCAAGACACAAGATCACAGAGACAG...         270.0
 4   GATGCTGGAATCACCCAGAGCCCAAGATACAAGATCACAGAGACAG...         270.0
 ..                                                ...           ...
 84  ATATCTGGAGTCTCCCACAACCCCAGACACAAGATCACAAAGAGGG...         273.0
 85  GATACTGGAGTCTCCCAGAACCCCAGACACAAGATCACAAAGAGGG...         273.0
 86  GATACTGGAGTCTCCCAGAACCCCAGACACAAGATCACAAAGAGGG...         273.0
 87  CACAACCGCCTTTATTGGTACCGACAGACCCTGGGGCAGGGCCCAG...         189.0
 88  GATTCTGGAGTCACACAAACCCCAAAGCACCTGATCACAGCAACTG...         270.0

 [89 rows x 3 columns],
 'D':          name             value
 id
 0    TRBD1*01      GGGACAGGGGGC
 1    TRBD2*01  GGGACTAGCGGGGGGG
 2    TRBD2*02  GGGACTAGCGGGAGGG,
 'J':                                                  name  \
 id
 0   K02545|TRBJ1-1*01|Homo sapiens|F|J-REGION|749....
 1   K02545|TRBJ1-2*01|Homo sapiens|F|J-REGION|886....
 2   M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...
 3   M14158|TRBJ1-4*01|Homo sapiens|F|J-REGION|2095...
 4   M14158|TRBJ1-5*01|Homo sapiens|F|J-REGION|2368...
 5   M14158|TRBJ1-6*01|Homo sapiens|F|J-REGION|2859...
 6   L36092|TRBJ1-6*02|Homo sapiens|F|J-REGION|6430...
 7   X02987|TRBJ2-1*01|Homo sapiens|F|J-REGION|800....
 8   X02987|TRBJ2-2*01|Homo sapiens|F|J-REGION|995....
 9   X02987|TRBJ2-3*01|Homo sapiens|F|J-REGION|1282...
 10  X02987|TRBJ2-4*01|Homo sapiens|F|J-REGION|1432...
 11  X02987|TRBJ2-5*01|Homo sapiens|F|J-REGION|1553...
 12  X02987|TRBJ2-6*01|Homo sapiens|F|J-REGION|1673...
 13  M14159|TRBJ2-7*01|Homo sapiens|F|J-REGION|2316...
 14  X02987|TRBJ2-7*02|Homo sapiens|ORF|J-REGION|18...

                                                 value  anchor_index
 id
 0    TGAACACTGAAGCTTTCTTTGGACAAGGCACCAGACTCACAGTTGTAG            17
 1    CTAACTATGGCTACACCTTCGGTTCGGGGACCAGGTTAACCGTTGTAG            17
 2   CTCTGGAAACACCATATATTTTGGAGAGGGAAGTTGGCTCACTGTT...            19
 3   CAACTAATGAAAAACTGTTTTTTGGCAGTGGAACCCAGCTCTCTGT...            20
 4   TAGCAATCAGCCCCAGCATTTTGGTGATGGGACTCGACTCTCCATC...            19
 5   CTCCTATAATTCACCCCTCCACTTTGGGAATGGGACCAGGCTCACT...            22
 6   CTCCTATAATTCACCCCTCCACTTTGGGAACGGGACCAGGCTCACT...            22
 7   CTCCTACAATGAGCAGTTCTTCGGGCCAGGGACACGGCTCACCGTG...            19
 8   CGAACACCGGGGAGCTGTTTTTTGGAGAAGGCTCTAGGCTGACCGT...            20
 9   AGCACAGATACGCAGTATTTTGGCCCAGGCACCCGGCTGACAGTGCTCG            18
 10  AGCCAAAAACATTCAGTACTTCGGCGCCGGGACCCGGCTCTCAGTG...            19
 11   ACCAAGAGACCCAGTACTTCGGGCCAGGCACGCGGCTCCTGGTGCTCG            17
 12  CTCTGGGGCCAACGTCCTGACTTTCGGGGCCGGCAGCAGGCTGACC...            22
 13    CTCCTACGAGCAGTACTTCGGGCCGGGCACCAGGCTCACGGTCACAG            16
 14    CTCCTACGAGCAGTACGTCGGGCCGGGCACCAGGCTCACGGTCACAG            16  }

First we make a copy of genomic_dataframe_dict

import copy
genomic_dict = copy.deepcopy(mdl_hb.genomic_dataframe_dict)
genomic_dict['V']['name']

id
0     U66059|TRBV1*01|Homo sapiens|P|V-REGION|91723....
1     U66059|TRBV10-1*01|Homo sapiens|F|V-REGION|214...
2     AF009660|TRBV10-1*02|Homo sapiens|F|V-REGION|5...
3     AF009660|TRBV10-1*04|Homo sapiens|F|V-REGION|5...
4     U66059|TRBV10-2*01|Homo sapiens|F|V-REGION|239...
                            ...
84    M14261|TRBV7-9*04|Homo sapiens|(F)|V-REGION|58...
85    M27385|TRBV7-9*05|Homo sapiens|(F)|V-REGION|34...
86    X74844|TRBV7-9*06|Homo sapiens|(F)|V-REGION|10...
87    L14854|TRBV7-9*07|Homo sapiens|(F)|V-REGION|1....
88    U66059|TRBV9*01|Homo sapiens|F|V-REGION|206836...
Name: name, Length: 89, dtype: object

Then we change the long description names using v_genLabel function.

genomic_dict['V']['name'] = p3.v_genLabel(genomic_dict['V']['name'])
genomic_dict['J']['name'] = p3.v_genLabel(genomic_dict['J']['name'])

Finally, update the the genomic_dataframe_dict with this new dict, by using the set_genomic_dataframe_dict method.

mdl_hb.set_genomic_dataframe_dict(genomic_dict)

mdl_hb.parms['v_choice']

	value	name
id
0	GATACTGGAATTACCCAGACACCAAAATACCTGGTCACAGCAATGG...	TRBV1*01
1	GATGCTGAAATCACCCAGAGCCCAAGACACAAGATCACAGAGACAG...	TRBV10-1*01
2	GATGCTGAAATCACCCAGAGCCCAAGACACAAGATCACAGAGACAG...	TRBV10-1*02
3	GATGCTGAAATCACCCAGAGCCCAAGACACAAGATCACAGAGACAG...	TRBV10-1*04
4	GATGCTGGAATCACCCAGAGCCCAAGATACAAGATCACAGAGACAG...	TRBV10-2*01
...	...	...
84	ATATCTGGAGTCTCCCACAACCCCAGACACAAGATCACAAAGAGGG...	TRBV7-9*04
85	GATACTGGAGTCTCCCAGAACCCCAGACACAAGATCACAAAGAGGG...	TRBV7-9*05
86	GATACTGGAGTCTCCCAGAACCCCAGACACAAGATCACAAAGAGGG...	TRBV7-9*06
87	CACAACCGCCTTTATTGGTACCGACAGACCCTGGGGCAGGGCCCAG...	TRBV7-9*07
88	GATTCTGGAGTCACACAAACCCCAAAGCACCTGATCACAGCAACTG...	TRBV9*01

89 rows × 2 columns

mdl_hb['v_choice']

<xarray.DataArray (v_choice: 89)>
array([4.88741e-03, 9.32369e-03, 9.32259e-03, 1.30320e-02, 3.43430e-04,
       8.50694e-03, 7.71250e-03, 6.12276e-04, 5.06104e-03, 4.59289e-05,
       4.48245e-03, 8.16181e-03, 7.00053e-04, 7.77164e-03, 1.16174e-02,
       1.16158e-02, 1.13872e-02, 1.06555e-02, 2.36870e-03, 2.28368e-02,
       1.56715e-04, 4.58447e-03, 0.00000e+00, 4.88782e-05, 0.00000e+00,
       1.54500e-02, 2.74050e-02, 5.18979e-03, 4.80289e-03, 1.62765e-01,
       6.61229e-02, 2.07174e-02, 7.36226e-04, 2.47846e-37, 1.07472e-02,
       2.03820e-02, 9.53245e-03, 9.53006e-03, 2.35607e-03, 2.39208e-03,
       2.39208e-03, 2.45219e-03, 1.66904e-02, 2.98807e-03, 2.98804e-03,
       1.59156e-02, 1.25468e-02, 9.06711e-03, 9.06936e-03, 9.76534e-02,
       9.57746e-03, 9.17011e-03, 1.14472e-02, 1.14458e-02, 9.77554e-03,
       1.89924e-02, 3.97949e-04, 1.70242e-03, 8.91336e-03, 6.73926e-03,
       6.69167e-03, 2.94257e-02, 1.38534e-02, 4.19227e-03, 4.19275e-03,
       2.48630e-03, 4.83399e-04, 1.29795e-07, 6.49975e-05, 1.94736e-02,
       1.94728e-02, 0.00000e+00, 3.91595e-03, 3.96230e-02, 0.00000e+00,
       0.00000e+00, 6.73805e-05, 2.90073e-02, 9.28490e-03, 5.32118e-03,
       8.18432e-03, 1.07035e-03, 3.64764e-04, 3.26790e-03, 3.26801e-03,
       3.26801e-03, 3.26809e-03, 8.23600e-04, 1.56399e-02])
Coordinates:
  * v_choice       (v_choice) int64 0 1 2 3 4 5 6 7 ... 81 82 83 84 85 86 87 88
    lbl__v_choice  (v_choice) object 'TRBV1*01' 'TRBV10-1*01' ... 'TRBV9*01'
    seq__v_choice  (v_choice) object 'GATACTGGAATTACCCAGACACCAAAATACCTGGTCACA...
Attributes:
    nickname:    v_choice
    event_type:  GeneChoice
    seq_type:    V_gene
    seq_side:    Undefined_side
    priority:    7
    parents:     []
    childs:      ['v_3_del', 'd_gene', 'j_choice']

mdl_hb.ErrorRate_dict

{'error_type': 'SingleErrorRate', 'error_values': '0.000396072'}

New models

Pygor has a methods to create default VDJ and VJ models from a dataframes.

new_V_gene_dict = {
    'name':'my_pseudo_TRBV',
    'value':'AAACCCTTTGGGACCCAGAGCCCAAGACACAAGATCACAGAGACAGGAAGGCAGGTGACCTTGGCGTGTCACCAGACTTGGAACCACAACAATATGTTCTGGTATCGACAAGACCTGGGACATGGGCTGAGGCTGATCCATTACTCATATGGTGTTCACGACACTAACAAAGGAGAAGTCTCAGATGGCTACAGTGTCTCTAGATCAAACACAGAGGACCTCCCCCTCACTCTGTAGTCTGCTGCCTCCTCCCAGACATCTGTATATTTCTGCGCCAGCAGTGAGTC',
    'anchor_index': 270
}

df_V = genomic_dict['V'].loc[10:15]
df_V = df_V.append(new_V_gene_dict, ignore_index=True)
df_V.index.name='id'
df_V

	name	value	anchor_index
id
0	TRBV11-2*03	GAAGCTGGAGTTGCCCAGTCTCCCAGATATAAGATTATAGAGAAAA...	273.0
1	TRBV11-3*01	GAAGCTGGAGTGGTTCAGTCTCCCAGATATAAGATTATAGAGAAAA...	273.0
2	TRBV12-1*01	GCTGGTGTTATCCAGTCACCCAGGCACAAAGTGACAGAGATGGGAC...	270.0
3	TRBV12-2*01	GCTGGCATTATCCAGTCACCCAAGCATGAGGTGACAGAAATGGGAC...	270.0
4	TRBV12-3*01	GATGCTGGAGTTATCCAGTCACCCCGCCATGAGGTGACAGAGATGG...	273.0
5	TRBV12-4*01	GATGCTGGAGTTATCCAGTCACCCCGGCACGAGGTGACAGAGATGG...	273.0
6	my_pseudo_TRBV	AAACCCTTTGGGACCCAGAGCCCAAGACACAAGATCACAGAGACAG...	270.0

Now we can use this new genomic templates to create a new model

new_mdl_0 = p3.IgorModel.make_default_VDJ(df_V, genomic_dict['D'], genomic_dict['J'])
new_mdl_0

<pygor3.IgorIO.IgorModel at 0x7fac17fdd490>

new_mdl_0.plot_Event('v_choice')

(<Figure size 1296x1080 with 1 Axes>,
 <AxesSubplot:title={'center':'$P($v_choice$)$'}>)

By default the new model will initiate with a uniform probability, that can be used to infer a new model from data

Exporting a Model

Conditional and marginal probabilities can be exported as pdf files with plots

fln_output_prefix = "mdl_hb"
mdl_hb.export_plot_Pconditionals(fln_output_prefix+"_CP")

mdl_hb.export_plot_Pmarginals(fln_output_prefix+"_MP")

A model can be exported in IGoR’s format with write_model method

new_mdl_0.write_model('new_model_parms.txt', 'new_model_marginals.txt',
                      fln_V_gene_CDR3_anchors='new_V_anchors.csv',
                      fln_J_gene_CDR3_anchors='new_J_anchors.csv')

Writing model parms in file  new_model_parms.txt
Writing model marginals in file  new_model_marginals.txt
Writing gene anchor's in file  new_V_anchors.csv
Writing gene anchor's in file  new_J_anchors.csv

!head new_model_parms.txt

@Event_list
#GeneChoice;V_gene;Undefined_side;7;v_choice
%TRBV11-2*03;GAAGCTGGAGTTGCCCAGTCTCCCAGATATAAGATTATAGAGAAAAGGCAGAGTGTGGCTTTTTGGTGCAATCCTATATCTGGCCATGCTACCCTTTACTGGTACCAGCAGATCCTGGGACAGGGCCCAAAGCTTCTGATTCAGTTTCAGAATAACGGTGTAGTGGATGATTCACAGTTGCCTAAGGATCGATTTTCTGCAGAGAGGCTCAAAGGAGTAGACTCCACTCTCAAGATCCAACCTGCAAAGCTTGAGGACTCGGCCGTGTATCTCTGTGCCAGCAGCTTAGA;0
%TRBV11-3*01;GAAGCTGGAGTGGTTCAGTCTCCCAGATATAAGATTATAGAGAAAAAACAGCCTGTGGCTTTTTGGTGCAATCCTATTTCTGGCCACAATACCCTTTACTGGTACCTGCAGAACTTGGGACAGGGCCCGGAGCTTCTGATTCGATATGAGAATGAGGAAGCAGTAGACGATTCACAGTTGCCTAAGGATCGATTTTCTGCAGAGAGGCTCAAAGGAGTAGACTCCACTCTCAAGATCCAGCCTGCAGAGCTTGGGGACTCGGCCGTGTATCTCTGTGCCAGCAGCTTAGA;1
%TRBV12-1*01;GCTGGTGTTATCCAGTCACCCAGGCACAAAGTGACAGAGATGGGACAATCAGTAACTCTGAGATGCGAACCAATTTCAGGCCACAATGATCTTCTCTGGTACAGACAGACCTTTGTGCAGGGACTGGAATTGCTGAATTACTTCTGCAGCTGGACCCTCGTAGATGACTCAGGAGTGTCCAAGGATTGATTCTCAGCACAGATGCCTGATGTATCATTCTCCACTCTGAGGATCCAGCCCATGGAACCCAGGGACTTGGGCCTATATTTCTGTGCCAGCAGCTTTGC;2
%TRBV12-2*01;GCTGGCATTATCCAGTCACCCAAGCATGAGGTGACAGAAATGGGACAAACAGTGACTCTGAGATGTGAGCCAATTTTTGGCCACAATTTCCTTTTCTGGTACAGAGATACCTTCGTGCAGGGACTGGAATTGCTGAGTTACTTCCGGAGCTGATCTATTATAGATAATGCAGGTATGCCCACAGAGCGATTCTCAGCTGAGAGGCCTGATGGATCATTCTCTACTCTGAAGATCCAGCCTGCAGAGCAGGGGGACTCGGCCGTGTATGTCTGTGCAAGTCGCTTAGC;3
%TRBV12-3*01;GATGCTGGAGTTATCCAGTCACCCCGCCATGAGGTGACAGAGATGGGACAAGAAGTGACTCTGAGATGTAAACCAATTTCAGGCCACAACTCCCTTTTCTGGTACAGACAGACCATGATGCGGGGACTGGAGTTGCTCATTTACTTTAACAACAACGTTCCGATAGATGATTCAGGGATGCCCGAGGATCGATTCTCAGCTAAGATGCCTAATGCATCATTCTCCACTCTGAAGATCCAGCCCTCAGAACCCAGGGACTCAGCTGTGTACTTCTGTGCCAGCAGTTTAGC;4
%TRBV12-4*01;GATGCTGGAGTTATCCAGTCACCCCGGCACGAGGTGACAGAGATGGGACAAGAAGTGACTCTGAGATGTAAACCAATTTCAGGACACGACTACCTTTTCTGGTACAGACAGACCATGATGCGGGGACTGGAGTTGCTCATTTACTTTAACAACAACGTTCCGATAGATGATTCAGGGATGCCCGAGGATCGATTCTCAGCTAAGATGCCTAATGCATCATTCTCCACTCTGAAGATCCAGCCCTCAGAACCCAGGGACTCAGCTGTGTACTTCTGTGCCAGCAGTTTAGC;5
%my_pseudo_TRBV;AAACCCTTTGGGACCCAGAGCCCAAGACACAAGATCACAGAGACAGGAAGGCAGGTGACCTTGGCGTGTCACCAGACTTGGAACCACAACAATATGTTCTGGTATCGACAAGACCTGGGACATGGGCTGAGGCTGATCCATTACTCATATGGTGTTCACGACACTAACAAAGGAGAAGTCTCAGATGGCTACAGTGTCTCTAGATCAAACACAGAGGACCTCCCCCTCACTCTGTAGTCTGCTGCCTCCTCCCAGACATCTGTATATTTCTGCGCCAGCAGTGAGTC;6
#GeneChoice;J_gene;Undefined_side;7;j_choice

!head new_V_anchors.csv

gene;anchor_index
TRBV11-2*03;273
TRBV11-3*01;273
TRBV12-1*01;270
TRBV12-2*01;270
TRBV12-3*01;273
TRBV12-4*01;273
my_pseudo_TRBV;270

mdl_0 = p3.IgorModel('new_model_parms.txt', 'new_model_marginals.txt',
                      fln_V_gene_CDR3_anchors='new_V_anchors.csv',
                      fln_J_gene_CDR3_anchors='new_J_anchors.csv')
mdl_0['v_choice']

Reading Parms filename from:  new_model_parms.txt
Reading Marginals filename from:  new_model_marginals.txt

<xarray.DataArray (v_choice: 7)>
array([0.14285714, 0.14285714, 0.14285714, 0.14285714, 0.14285714,
       0.14285714, 0.14285714])
Coordinates:
  * v_choice       (v_choice) int64 0 1 2 3 4 5 6
    lbl__v_choice  (v_choice) object 'TRBV11-2*03' ... 'my_pseudo_TRBV'
    seq__v_choice  (v_choice) object 'GAAGCTGGAGTTGCCCAGTCTCCCAGATATAAGATTATA...
Attributes:
    nickname:    v_choice
    event_type:  GeneChoice
    seq_type:    V_gene
    seq_side:    Undefined_side
    priority:    7
    parents:     []
    childs:      ['j_choice', 'd_gene', 'v_3_del']

IgorModel can also be exported in separated csv files

mdl_0.export_csv('initial_')

!head initial_P__insertions.csv

Insertions;P(vd_ins);P(dj_ins)
0;0.024390243902439025;0.024390243902439025
1;0.024390243902439025;0.024390243902439025
2;0.024390243902439025;0.024390243902439025
3;0.024390243902439025;0.024390243902439025
4;0.024390243902439025;0.024390243902439025
5;0.024390243902439025;0.024390243902439025
6;0.024390243902439025;0.024390243902439025
7;0.024390243902439025;0.024390243902439025
8;0.024390243902439025;0.024390243902439025

!head initial_P__j_choice__G__v_choice.csv

;TRBJ1-1*01;TRBJ1-2*01;TRBJ1-3*01;TRBJ1-4*01;TRBJ1-5*01;TRBJ1-6*01;TRBJ1-6*02;TRBJ2-1*01;TRBJ2-2*01;TRBJ2-3*01;TRBJ2-4*01;TRBJ2-5*01;TRBJ2-6*01;TRBJ2-7*01;TRBJ2-7*02
TRBV11-2*03;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667
TRBV11-3*01;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667
TRBV12-1*01;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667
TRBV12-2*01;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667
TRBV12-3*01;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667
TRBV12-4*01;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667
my_pseudo_TRBV;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667;0.06666666666666667

Entropy

$ H = - \sumx P(x) :nbsphinx-math:`log`(P(x)) $

import numpy as np
-np.dot(mdl_hb['v_choice'], np.nan_to_num(np.log2(mdl_hb['v_choice']), neginf=0, nan=0) )

/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)

5.252905287497762

mdl_hb.get_entropy_event('v_choice')

<xarray.DataArray ()>
array(5.25291424)

Mutual Information

I_V_J = mdl_hb.get_mutual_information_events('v_choice', 'j_choice')
I_V_J

/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)

<xarray.DataArray ()>
array(0.0779008)

da_mi = mdl_hb.get_mutual_information()
da_mi

/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)

<xarray.DataArray 'mutual_information' (x: 9, y: 9)>
array([[0.00000000e+00, 2.74757454e-02, 7.79007955e-02, 3.34244341e-01,
        2.42224927e-03, 3.84618632e-03, 3.11240332e-03, 0.00000000e+00,
        0.00000000e+00],
       [2.74757454e-02, 0.00000000e+00, 1.54509182e-01, 7.40373019e-04,
        1.50139212e-01, 2.66774661e-01, 4.92915719e-03, 0.00000000e+00,
        0.00000000e+00],
       [7.79007955e-02, 1.54509182e-01, 0.00000000e+00, 7.48363481e-04,
        9.41603438e-03, 8.72144895e-03, 3.15006638e-01, 0.00000000e+00,
        0.00000000e+00],
       [3.34244341e-01, 7.40373019e-04, 7.48363481e-04, 0.00000000e+00,
        5.45436417e-05, 7.33990325e-05, 2.35845118e-05, 0.00000000e+00,
        0.00000000e+00],
       [2.42224927e-03, 1.50139212e-01, 9.41603438e-03, 5.45436417e-05,
        0.00000000e+00, 4.49012987e-01, 2.87003080e-04, 0.00000000e+00,
        0.00000000e+00],
       [3.84618632e-03, 2.66774661e-01, 8.72144895e-03, 7.33990325e-05,
        4.49012987e-01, 0.00000000e+00, 2.74201841e-04, 0.00000000e+00,
        0.00000000e+00],
       [3.11240332e-03, 4.92915719e-03, 3.15006638e-01, 2.35845118e-05,
        2.87003080e-04, 2.74201841e-04, 0.00000000e+00, 0.00000000e+00,
        0.00000000e+00],
       [0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
        0.00000000e+00],
       [0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
        0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
        0.00000000e+00]])
Coordinates:
  * x        (x) <U8 'v_choice' 'd_gene' 'j_choice' ... 'vd_ins' 'dj_ins'
  * y        (y) <U8 'v_choice' 'd_gene' 'j_choice' ... 'vd_ins' 'dj_ins'

mdl_hb.plot_mutual_information(da_mi)

<AxesSubplot:>

event_nickname1 = 'v_choice'
event_nickname2 = 'j_choice'

mdl = p3.get_default_IgorModel("human", "tcr_beta")

da_P_x_y = mdl.get_P_joint([event_nickname1, event_nickname2])
da_P_x = mdl.Pmarginal[event_nickname1]
da_P_y = mdl.Pmarginal[event_nickname2]

da_P_x_times_P_y = (da_P_x*da_P_y)
da_P_x_times_P_y

da_log_P_ratio = xr.zeros_like(da_P_x_y)

da_log_P_ratio.values = np.nan_to_num(
    np.log2(da_P_x_y / da_P_x_times_P_y), nan=0.0, neginf=0.0
)

# da_log_Value.values = np_log_Value
xr.dot( da_P_x_y, da_log_P_ratio )

Reading Parms filename from:  /home/olivares/.local/share/igor/models/human/tcr_beta/models/model_parms.txt
Reading Marginals filename from:  /home/olivares/.local/share/igor/models/human/tcr_beta/models/model_marginals.txt

/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)

<xarray.DataArray ()>
array(0.0779008)

da_mi = mdl_hb.get_mutual_information()

/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)

mdl_hb.plot_mutual_information(da_mi)

<AxesSubplot:>

da_mi = mdl_hb.get_mutual_information()

/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)
/home/olivares/anaconda3/envs/statbiophys-dev/lib/python3.7/site-packages/xarray/core/computation.py:739: RuntimeWarning: divide by zero encountered in log2
  result_data = func(*input_data)

mdl_hb.plot_mutual_information(da_mi)

<AxesSubplot:>

As can be seen in the Bayesian network, the number of deletions in V depends on choosen V

Select events by probability

Here we can see which combinations of V and D are not possible, for our model

Pjoint_V_D = mdl_hb.get_P_joint(['v_choice', 'd_gene'])

da_tmp = Pjoint_V_D.where(Pjoint_V_D == 0)

df = da_tmp.to_dataframe('P_joint_V_D').dropna()
df #.to_csv('bibibi.csv', sep=';')

		lbl__v_choice	seq__v_choice	lbl__d_gene	seq__d_gene	P_joint_V_D
v_choice	d_gene
22	0	TRBV15*03	GATGCCATGGTCATCCAGAACCCAAGATACCGGGTTACCCAGTTTG...	TRBD1*01	GGGACAGGGGGC	0.0
	1	TRBV15*03	GATGCCATGGTCATCCAGAACCCAAGATACCGGGTTACCCAGTTTG...	TRBD2*01	GGGACTAGCGGGGGGG	0.0
	2	TRBV15*03	GATGCCATGGTCATCCAGAACCCAAGATACCGGGTTACCCAGTTTG...	TRBD2*02	GGGACTAGCGGGAGGG	0.0
24	0	TRBV17*01	GAGCCTGGAGTCAGCCAGACCCCCAGACACAAGGTCACCAACATGG...	TRBD1*01	GGGACAGGGGGC	0.0
	1	TRBV17*01	GAGCCTGGAGTCAGCCAGACCCCCAGACACAAGGTCACCAACATGG...	TRBD2*01	GGGACTAGCGGGGGGG	0.0
	2	TRBV17*01	GAGCCTGGAGTCAGCCAGACCCCCAGACACAAGGTCACCAACATGG...	TRBD2*02	GGGACTAGCGGGAGGG	0.0
33	1	TRBV26*01	GATGCTGTAGTTACACAATTCCCAAGACACAGAATCATTGGGACAG...	TRBD2*01	GGGACTAGCGGGGGGG	0.0
	2	TRBV26*01	GATGCTGTAGTTACACAATTCCCAAGACACAGAATCATTGGGACAG...	TRBD2*02	GGGACTAGCGGGAGGG	0.0
67	0	TRBV6-9*01	AATGCTGGTGTCACTCAGACCCCAAAATTCCACATCCTGAAGACAG...	TRBD1*01	GGGACAGGGGGC	0.0
	2	TRBV6-9*01	AATGCTGGTGTCACTCAGACCCCAAAATTCCACATCCTGAAGACAG...	TRBD2*02	GGGACTAGCGGGAGGG	0.0
71	0	TRBV7-2*03	GCTGGAGTCTCCCAGTCCCCCAGTAACAAGGTCACAGAGAAGGGAA...	TRBD1*01	GGGACAGGGGGC	0.0
	1	TRBV7-2*03	GCTGGAGTCTCCCAGTCCCCCAGTAACAAGGTCACAGAGAAGGGAA...	TRBD2*01	GGGACTAGCGGGGGGG	0.0
	2	TRBV7-2*03	GCTGGAGTCTCCCAGTCCCCCAGTAACAAGGTCACAGAGAAGGGAA...	TRBD2*02	GGGACTAGCGGGAGGG	0.0
74	0	TRBV7-3*02	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...	TRBD1*01	GGGACAGGGGGC	0.0
	1	TRBV7-3*02	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...	TRBD2*01	GGGACTAGCGGGGGGG	0.0
	2	TRBV7-3*02	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...	TRBD2*02	GGGACTAGCGGGAGGG	0.0
75	0	TRBV7-3*03	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...	TRBD1*01	GGGACAGGGGGC	0.0
	1	TRBV7-3*03	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...	TRBD2*01	GGGACTAGCGGGGGGG	0.0
	2	TRBV7-3*03	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...	TRBD2*02	GGGACTAGCGGGAGGG	0.0

mdl_hb['v_choice'].loc[22] #, mdl_hb['v_choice'].loc[67]

<xarray.DataArray ()>
array(0.)
Coordinates:
    v_choice       int64 22
    lbl__v_choice  object 'TRBV15*03'
    seq__v_choice  object 'GATGCCATGGTCATCCAGAACCCAAGATACCGGGTTACCCAGTTTGGAAA...
Attributes:
    nickname:    v_choice
    event_type:  GeneChoice
    seq_type:    V_gene
    seq_side:    Undefined_side
    priority:    7
    parents:     []
    childs:      ['v_3_del', 'd_gene', 'j_choice']

mdl_hb.genomic_dataframe_dict['V'].loc[22]

name                                                    TRBV15*03
value           GATGCCATGGTCATCCAGAACCCAAGATACCGGGTTACCCAGTTTG...
anchor_index                                                270.0
Name: 22, dtype: object

Use a default IGoR’s model to generate sequences

df_seqs = p3.generate(10, mdl_hb)
df_seqs

Writing model parms in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata/models/model_parms.txt
Writing model marginals in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata/models/model_marginals.txt
Writing gene anchor's in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata/ref_genome/V_gene_CDR3_anchors.csv
Writing gene anchor's in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata/ref_genome/J_gene_CDR3_anchors.csv
Writing model parms in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//models/model_parms.txt
Writing model marginals in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//models/model_marginals.txt
Writing gene anchor's in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//ref_genome/V_gene_CDR3_anchors.csv
Writing gene anchor's in file  ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//ref_genome/J_gene_CDR3_anchors.csv
/home/olivares/.local/bin/igor -set_wd ./igor_generating_3z_y24jv -batch dataIGoRBTnea5JaoG -set_custom_model ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//models/model_parms.txt ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//models/model_marginals.txt -set_CDR3_anchors  --V ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//ref_genome/V_gene_CDR3_anchors.csv --J ./igor_generating_3z_y24jv/dataIGoRBTnea5JaoG_mdldata//ref_genome/J_gene_CDR3_anchors.csv -generate 10

	nt_sequence
seq_index
0	AAGGCTGGAGTCACTCAAACTCCAAGATATCTGATCAAAACGAGAG...
1	GAAGCTGGAGTTACTCAGTTCCCCAGCCACAGCGTAATAGAGAAGG...
2	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...
3	GGAGCTGGAGTCTCCCAGTCCCCCAGTAACAAGGTCACAGAGAAGG...
4	GAAGCTGGAGTTGCCCAGTCTCCCAGATATAAGATTATAGAGAAAA...
5	GGTGCTGGAGTCTCCCAGACCCCCAGTAACAAGGTCACAGAGAAGG...
6	GATGCTGATGTTACCCAGACCCCAAGGAATAGGATCACAAAGACAG...
7	AAGGCTGGAGTCACTCAAACTCCAAGATATCTGATCAAAACGAGAG...
8	GAAACGGGAGTTACGCAGACACCAAGACACCTGGTCATGGGAATGA...
9	CATGCCAAAGTCACACAGACTCCAGGACATTTGGTCAAAGGAAAAG...

One sequence evaluation

Let’s consider that we have a sequence of a TCR \(\beta\) receptor

str_seq_hb = "ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGTATCGTCAGTTCCCGAAACAGAGTCTCATGCTGATGGCAACTTCCAATGAGGGCTCCAAGGCCACATACGAGCAAGGCGTCGAGAAGGACAAGTTTCTCATCAACCATGCAAGCCTGACCTTGTCCACTCTGACAGTGACCAGTGCCCATCCTGAAGACAGCAGCTTCTACATCTGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTTGGAG"

A classical approach would be make alignments of the VDJ segments and consider the maximun alignment of the segments as uniquely determinated construction.

## TODO: SHOW NAIVE ALIGNMENT

## TODO: Simple explantion of the inference process

Evaluate VDJ model

mdl = p3.get_default_IgorModel("human", "tcr_beta")

Reading Parms filename from:  /home/olivares/.local/share/igor/models/human/tcr_beta/models/model_parms.txt
Reading Marginals filename from:  /home/olivares/.local/share/igor/models/human/tcr_beta/models/model_marginals.txt

help(p3.evaluate)

Help on function evaluate in module pygor3.IgorIO:

evaluate(input_sequences: Union[str, pandas.core.frame.DataFrame, numpy.ndarray, pathlib.Path], mdl: pygor3.IgorIO.IgorModel, N_scenarios=None, igor_wd=None, airr_format=True, batch_clean=True)
    Evaluate input sequences with provided model
    :param input_sequences:Union[str, pd.DataFrame, np.ndarray, Path]
    :param mdl:IgorModel
    :param batch_clean: Remove all temporary files True by default.

df_scenarios = p3.evaluate(str_seq_hb, mdl) # , N_scenarios=20
df_scenarios

Writing model parms in file  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/models/model_parms.txt
Writing model marginals in file  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/models/model_marginals.txt
Writing gene anchor's in file  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/V_gene_CDR3_anchors.csv
Writing gene anchor's in file  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/J_gene_CDR3_anchors.csv
/home/olivares/.local/bin/igor -set_wd ./igor_evaluating_1b6e6i7z -batch dataIGoRyNbbuLilFX -read_seqs ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFXinput_sequences.csv
/home/olivares/.local/bin/igor -set_wd ./igor_evaluating_1b6e6i7z -batch dataIGoRyNbbuLilFX -set_genomic  --V ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/genomicVs.fasta --D ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/genomicDs.fasta --J ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/genomicJs.fasta -set_CDR3_anchors  --V ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/V_gene_CDR3_anchors.csv --J ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/J_gene_CDR3_anchors.csv -align  --all
/home/olivares/.local/bin/igor -set_wd ./igor_evaluating_1b6e6i7z -batch dataIGoRyNbbuLilFX -set_custom_model ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/models/model_parms.txt ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/models/model_marginals.txt -evaluate  -output  --scenarios 10  --Pgen
./igor_evaluating_1b6e6i7z/aligns/dataIGoRyNbbuLilFX_indexed_CDR3s.csv
('',)
Incorrect number of bindings supplied. The current statement uses 5, and there are 1 supplied.
Loading Gene templates ...
V ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/genomicVs.fasta
J ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/genomicJs.fasta
D ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/genomicDs.fasta
loading Anchors data ...
Loading Gene Anchors from  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/V_gene_CDR3_anchors.csv
Loading Gene Anchors from  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/ref_genome/J_gene_CDR3_anchors.csv
./igor_evaluating_1b6e6i7z/aligns/dataIGoRyNbbuLilFX_V_alignments.csv
['']
['']
['']
Alignments loaded in database in ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX.db
Reading Parms filename from:  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/models/model_parms.txt
Reading Marginals filename from:  ./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_mdldata/models/model_marginals.txt
Loading parms to database
('v_choice', 'v_3_del')
('v_choice', 'd_gene')
('v_choice', 'j_choice')
('d_gene', 'd_3_del')
('d_gene', 'd_5_del')
('j_choice', 'j_5_del')
('j_choice', 'd_gene')
('d_5_del', 'd_3_del')
Loading marginals to database
./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_output/best_scenarios_counts.csv
events_name_nickname_dict {'GeneChoice_V_gene_Undefined_side_prio7_size89': 'v_choice', 'GeneChoice_J_gene_Undefined_side_prio7_size15': 'j_choice', 'GeneChoice_D_gene_Undefined_side_prio6_size3': 'd_gene', 'Deletion_V_gene_Three_prime_prio5_size21': 'v_3_del', 'Deletion_D_gene_Five_prime_prio5_size21': 'd_5_del', 'Deletion_J_gene_Five_prime_prio5_size23': 'j_5_del', 'Deletion_D_gene_Three_prime_prio5_size21': 'd_3_del', 'Insertion_VD_genes_Undefined_side_prio4_size31': 'vd_ins', 'DinucMarkov_VD_genes_Undefined_side_prio3_size16': 'vd_dinucl', 'Insertion_DJ_gene_Undefined_side_prio2_size31': 'dj_ins', 'DinucMarkov_DJ_gene_Undefined_side_prio1_size16': 'dj_dinucl'}
./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_output/Pgen_counts.csv
./igor_evaluating_1b6e6i7z/dataIGoRyNbbuLilFX_output/Pgen_counts.csv
('',)
Incorrect number of bindings supplied. The current statement uses 2, and there are 1 supplied.
----- Marginals -----
d_3_del
d_5_del
d_gene
dj_dinucl
dj_ins
j_5_del
j_choice
v_3_del
v_choice
vd_dinucl
vd_ins

	sequence_id	sequence	rev_comp	productive	v_call	d_call	j_call	sequence_alignment	germline_alignment	junction	junction_aa	v_cigar	d_cigar	j_cigar	v_score	v_identity	v_support	v_sequence_start	v_sequence_end	v_germline_start	v_germline_end	v_alignment_start	v_alignment_end	d_score	d_identity	d_support	d_sequence_start	d_sequence_end	d_germline_start	d_germline_end	d_alignment_start	d_alignment_end	j_score	j_identity	j_support	j_sequence_start	j_sequence_end	j_germline_start	j_germline_end	j_alignment_start	j_alignment_end	sequence_aa	vj_in_frame	stop_codon	complete_vdj	locus	sequence_alignment_aa	n1_length	np1	np1_aa	np1_length	n2_length	np2	np2_aa	np2_length	p3v_length	p5d_length	p3d_length	p5j_length	scenario_rank	scenario_proba_cond_seq	pgen	quality	quality_alignment
0	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD1*01	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	4M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	20	NaN	NaN	292	294	7	10	NaN	NaN	240	NaN	NaN	7	53	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	5	GGGGA	NaN	5	4	CAGC	NaN	4	0	0	0	0	1	0.146984	4.871310e-13	NaN	NaN
1	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD2*02	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	4M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	20	NaN	NaN	292	294	14	17	NaN	NaN	240	NaN	NaN	7	53	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	5	GGGGA	NaN	5	4	CAGC	NaN	4	0	0	0	0	2	0.112527	4.871310e-13	NaN	NaN
2	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD1*01	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	3M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	15	NaN	NaN	296	297	6	8	NaN	NaN	240	NaN	NaN	3	49	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	9	GGGGAAGGG	NaN	9	1	C	NaN	1	0	0	0	0	3	0.062955	4.871310e-13	NaN	NaN
3	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD2*02	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	3M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	15	NaN	NaN	293	294	15	17	NaN	NaN	240	NaN	NaN	6	52	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	6	GGGGAA	NaN	6	3	AGC	NaN	4	0	0	1	0	4	0.060409	4.871310e-13	NaN	NaN
4	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD1*01	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	4M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	20	NaN	NaN	293	295	10	13	NaN	NaN	240	NaN	NaN	6	52	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	6	GGGGAA	NaN	6	3	AGC	NaN	3	0	0	0	0	5	0.034684	4.871310e-13	NaN	NaN
5	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD1*01	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	287M	4M	48M	1435	NaN	NaN	2	287	61	285	NaN	NaN	20	NaN	NaN	292	294	7	10	NaN	NaN	240	NaN	NaN	7	53	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	6	TGGGGA	NaN	6	4	CAGC	NaN	4	0	0	0	0	6	0.028806	4.871310e-13	NaN	NaN
6	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD2*02	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	3M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	15	NaN	NaN	293	294	11	13	NaN	NaN	240	NaN	NaN	6	52	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	6	GGGGAA	NaN	6	4	CAGC	NaN	4	0	0	0	0	7	0.023637	4.871310e-13	NaN	NaN
7	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD2*02	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	287M	4M	48M	1435	NaN	NaN	2	287	61	285	NaN	NaN	20	NaN	NaN	292	294	14	17	NaN	NaN	240	NaN	NaN	7	53	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	6	TGGGGA	NaN	6	4	CAGC	NaN	4	0	0	0	0	8	0.022053	4.871310e-13	NaN	NaN
8	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD2*02	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	4M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	20	NaN	NaN	288	290	11	14	NaN	NaN	240	NaN	NaN	11	57	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	1	G	NaN	1	8	AGGGCAGC	NaN	8	0	0	0	0	9	0.017562	4.871310e-13	NaN	NaN
9	0	ATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGT...	F	NaN	M11955|TRBV20-1*01|Homo sapiens|F|V-REGION|427...	TRBD2*02	M14158|TRBJ1-3*01|Homo sapiens|F|J-REGION|1499...	GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTG...	NaN	TGCAGTGCTGGGGAAGGGCAGCCTGGAAACACCATATATTTT	CSAGEGQPGNTIYF	288M	3M	48M	1440	NaN	NaN	2	288	61	286	NaN	NaN	15	NaN	NaN	292	293	14	16	NaN	NaN	240	NaN	NaN	7	53	4	51	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	5	GGGGA	NaN	5	5	GCAGC	NaN	5	0	0	0	0	10	0.015878	4.871310e-13	NaN	NaN

df_scenarios = p3.evaluate(str_seq_hb, mdl_hb, airr_format=False) # , N_scenarios=20
df_scenarios

Writing model parms in file  ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/models/model_parms.txt
Writing model marginals in file  ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/models/model_marginals.txt
Writing gene anchor's in file  ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/V_gene_CDR3_anchors.csv
Writing gene anchor's in file  ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/J_gene_CDR3_anchors.csv
/home/olivares/.local/bin/igor -set_wd ./igor_evaluating_mdsvldqo -batch dataIGoRQVzl4XnKII -read_seqs ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKIIinput_sequences.csv
/home/olivares/.local/bin/igor -set_wd ./igor_evaluating_mdsvldqo -batch dataIGoRQVzl4XnKII -set_genomic  --V ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/genomicVs.fasta --D ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/genomicDs.fasta --J ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/genomicJs.fasta -set_CDR3_anchors  --V ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/V_gene_CDR3_anchors.csv --J ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/ref_genome/J_gene_CDR3_anchors.csv -align  --all
/home/olivares/.local/bin/igor -set_wd ./igor_evaluating_mdsvldqo -batch dataIGoRQVzl4XnKII -set_custom_model ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/models/model_parms.txt ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_mdldata/models/model_marginals.txt -evaluate  -output  --scenarios 10  --Pgen
igor_fln_generated_realizations_werr:  ./igor_evaluating_mdsvldqo/dataIGoRQVzl4XnKII_output/best_scenarios_counts.csv

	scenario_rank	scenario_proba_cond_seq	v_choice	j_choice	d_gene	v_3_del	d_5_del	j_5_del	d_3_del	vd_ins	vd_dinucl	dj_ins	dj_dinucl	Mismatches	Pgen_estimate
seq_index
0	1	0.146984	29	2	0	9	9	6	7	5	[2, 2, 2, 2, 0]	4	[1, 2, 0, 1]	[]	4.871310e-13
0	2	0.112527	29	2	2	9	16	6	4	5	[2, 2, 2, 2, 0]	4	[1, 2, 0, 1]	[]	4.871310e-13
0	3	0.062955	29	2	0	9	8	6	9	9	[2, 2, 2, 2, 0, 0, 2, 2, 2]	1	[1]	[]	4.871310e-13
0	4	0.060409	29	2	2	9	17	6	3	6	[2, 2, 2, 2, 0, 0]	3	[1, 2, 0]	[]	4.871310e-13
0	5	0.034684	29	2	0	9	12	6	4	6	[2, 2, 2, 2, 0, 0]	3	[1, 2, 0]	[]	4.871310e-13
0	6	0.028806	29	2	0	10	9	6	7	6	[3, 2, 2, 2, 2, 0]	4	[1, 2, 0, 1]	[]	4.871310e-13
0	7	0.023637	29	2	2	9	13	6	8	6	[2, 2, 2, 2, 0, 0]	4	[1, 2, 0, 1]	[]	4.871310e-13
0	8	0.022053	29	2	2	10	16	6	4	6	[3, 2, 2, 2, 2, 0]	4	[1, 2, 0, 1]	[]	4.871310e-13
0	9	0.017562	29	2	2	9	13	6	7	1	[2]	8	[1, 2, 0, 1, 2, 2, 2, 0]	[]	4.871310e-13
0	10	0.015878	29	2	2	9	16	6	5	5	[2, 2, 2, 2, 0]	5	[1, 2, 0, 1, 2]	[]	4.871310e-13

Visualize scenarios

for ii, ps_scenario in df_scenarios.iterrows():
    mdl_hb.plot_scenario(ps_scenario) #, nt_lim=(200,340))

import pandas as pd
dict_scenario_fict = {
            'scenario_rank' : 1,
            'scenario_proba_cond_seq' : 0.9999,
            'v_choice' : 61,
            'j_choice' : 1,
            'd_gene' : 0,
            'v_3_del': 0,
            'd_5_del': 1,
            'j_5_del': 8,
            'd_3_del': 6,
            'vd_ins' : 1,
            'vd_dinucl': [1],
            'dj_ins': 2,
            'dj_dinucl': [0, 3],
            'Mismatches' : [],
            'norm_scenario_proba_cond_seq': 0.000225
        }
ps_scenario_fict = pd.Series(dict_scenario_fict)
mdl_hb.plot_scenario(ps_scenario_fict)

(<Figure size 1440x720 with 1 Axes>, <AxesSubplot:>)

df_scenario_aln_fict = mdl_hb.get_df_scenario_aln_from_scenario(ps_scenario_fict)
df_scenario_aln_fict
mdl_hb.get_gene_segment_dict('V', ps_scenario)

OrderedDict([('gene_template',
              'GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTGGAACCTCTGTGAAGATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGTATCGTCAGTTCCCGAAACAGAGTCTCATGCTGATGGCAACTTCCAATGAGGGCTCCAAGGCCACATACGAGCAAGGCGTCGAGAAGGACAAGTTTCTCATCAACCATGCAAGCCTGACCTTGTCCACTCTGACAGTGACCAGTGCCCATCCTGAAGACAGCAGCTTCTACATCTGCAGTGCTAGAGA'),
             ('int_gene_5_del', 0),
             ('int_gene_3_del', 5),
             ('palindrome_5_end', ''),
             ('gene_ini', 0),
             ('gene_end', 288),
             ('gene_cut',
              'GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTGGAACCTCTGTGAAGATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGTATCGTCAGTTCCCGAAACAGAGTCTCATGCTGATGGCAACTTCCAATGAGGGCTCCAAGGCCACATACGAGCAAGGCGTCGAGAAGGACAAGTTTCTCATCAACCATGCAAGCCTGACCTTGTCCACTCTGACAGTGACCAGTGCCCATCCTGAAGACAGCAGCTTCTACATCTGCAGTGCT'),
             ('palindrome_3_end', ''),
             ('gene_segment',
              'GGTGCTGTCGTCTCTCAACATCCGAGCTGGGTTATCTGTAAGAGTGGAACCTCTGTGAAGATCGAGTGCCGTTCCCTGGACTTTCAGGCCACAACTATGTTTTGGTATCGTCAGTTCCCGAAACAGAGTCTCATGCTGATGGCAACTTCCAATGAGGGCTCCAAGGCCACATACGAGCAAGGCGTCGAGAAGGACAAGTTTCTCATCAACCATGCAAGCCTGACCTTGTCCACTCTGACAGTGACCAGTGCCCATCCTGAAGACAGCAGCTTCTACATCTGCAGTGCT'),
             ('gene_description',
              "TRBV20-1*01 (v_choice: 29) (v_3_del: 9), 3'del : 5")])

mdl_hb.parms['d_gene']

	value	name
id
0	GGGACAGGGGGC	TRBD1*01
1	GGGACTAGCGGGGGGG	TRBD2*01
2	GGGACTAGCGGGAGGG	TRBD2*02

help(mdl_hb.plot_scenario)

Help on method plot_scenario in module pygor3.IgorIO:

plot_scenario(ps_scenario, nt_lim: Union[NoneType, tuple, list] = None, show_CDR3=True, ax=None) method of pygor3.IgorIO.IgorModel instance
    Return matplotlib fig, ax
    :param ps_scenario: Pandas Series scenario
    :param nt_lim:Union[None,tuple,list] region limits to show the scenario alignment
    default give boundaries around CDR3, if no anchors in model, show the whole scenario
    :param show_CDR3: Show CDR3 lines default(=True)

Observables from scenarios

Predefine Functions for Scenarios

fln_scenarios = "best_scenarios_counts.csv"
df_scenarios_many = mdl_hb.get_dataframe_scenarios(fln_scenarios)

igor_fln_generated_realizations_werr:  best_scenarios_counts.csv

Pairwise Probabilities

help(mdl_hb.get_P_from_scenarios_cols)

Help on method get_P_from_scenarios_cols in module pygor3.IgorIO:

get_P_from_scenarios_cols(df_scenarios, colname_list) method of pygor3.IgorIO.IgorModel instance
    Return xarray with marginalize probabilities of listed columns in dataframe scenarios df_scenarios
    :param df_scenarios: Scenarios with normalize probability. Loaded with self.get_dataframe_scenarios()
    :param colname_list: List of variables preserve for marginalization

Pjoint_V_J_scens = mdl_hb.get_P_from_scenarios_cols(df_scenarios_many, ['v_choice', 'j_choice'])
Pjoint_V_J_scens.plot(cmap='gnuplot2_r')

<matplotlib.collections.QuadMesh at 0x7fac14e44f10>

mdl_hb.get_P_joint(['v_choice', 'j_choice']).plot(cmap='gnuplot2_r')

<matplotlib.collections.QuadMesh at 0x7fac14d4ef50>

Mutual Information from scenarios

help(mdl_hb.get_mutual_information_events_from_df_scenarios)

Help on method get_mutual_information_events_from_df_scenarios in module pygor3.IgorIO:

get_mutual_information_events_from_df_scenarios(df_scenarios, event_nickname_x, event_nickname_y) method of pygor3.IgorIO.IgorModel instance
    Return mutual information in log10 of the desired events

mdl_hb.get_mutual_information_events_from_df_scenarios(df_scenarios_many, 'v_choice', 'j_choice')

mutual_information ( v_choice ,  j_choice ):  0.6392623620052954

<xarray.DataArray 'norm_scenario_proba_cond_seq' ()>
array(0.63926236)

da_mi_scenarios = mdl_hb.get_mutual_information_from_df_scenarios(df_scenarios_many)

mutual_information ( v_choice ,  j_choice ):  0.6392623620052954
mutual_information ( v_choice ,  d_gene ):  0.08894391386660068
mutual_information ( v_choice ,  v_3_del ):  0.7021801087511944
mutual_information ( v_choice ,  d_5_del ):  0.35886454309758736
mutual_information ( v_choice ,  j_5_del ):  0.4667910640736827
mutual_information ( v_choice ,  d_3_del ):  0.3381390129055559
mutual_information ( v_choice ,  vd_ins ):  0.36950632876861056
mutual_information ( v_choice ,  dj_ins ):  0.34050716113668833
mutual_information ( j_choice ,  d_gene ):  0.16839619808892387
mutual_information ( j_choice ,  v_3_del ):  0.1373839342864651
mutual_information ( j_choice ,  d_5_del ):  0.11353204598699257
mutual_information ( j_choice ,  j_5_del ):  0.44550937210770414
mutual_information ( j_choice ,  d_3_del ):  0.11443530326748617
mutual_information ( j_choice ,  vd_ins ):  0.11203857283894578
mutual_information ( j_choice ,  dj_ins ):  0.09776360260846355
mutual_information ( d_gene ,  v_3_del ):  0.017241089783731553
mutual_information ( d_gene ,  d_5_del ):  0.36172414085319887
mutual_information ( d_gene ,  j_5_del ):  0.021864784540005937
mutual_information ( d_gene ,  d_3_del ):  0.22353112818847967
mutual_information ( d_gene ,  vd_ins ):  0.013194437391113211
mutual_information ( d_gene ,  dj_ins ):  0.01294449164566809
mutual_information ( v_3_del ,  d_5_del ):  0.06598437525781697
mutual_information ( v_3_del ,  j_5_del ):  0.09843108497391649
mutual_information ( v_3_del ,  d_3_del ):  0.06941838791192181
mutual_information ( v_3_del ,  vd_ins ):  0.10582382685367009
mutual_information ( v_3_del ,  dj_ins ):  0.07374102134403618
mutual_information ( d_5_del ,  j_5_del ):  0.08378406617127063
mutual_information ( d_5_del ,  d_3_del ):  0.4950198124493378
mutual_information ( d_5_del ,  vd_ins ):  0.10780812230061945
mutual_information ( d_5_del ,  dj_ins ):  0.07617536406698441
mutual_information ( j_5_del ,  d_3_del ):  0.08559953020573469
mutual_information ( j_5_del ,  vd_ins ):  0.07299823195715721
mutual_information ( j_5_del ,  dj_ins ):  0.10957597446165623
mutual_information ( d_3_del ,  vd_ins ):  0.06822031197593291
mutual_information ( d_3_del ,  dj_ins ):  0.08640307827393552
mutual_information ( vd_ins ,  dj_ins ):  0.06747319489954777

da_mi_scenarios#.to_dataframe().dropna()

<xarray.DataArray 'mutual_information' (x: 9, y: 9)>
array([[0.        , 0.63926236, 0.08894391, 0.70218011, 0.35886454,
        0.46679106, 0.33813901, 0.36950633, 0.34050716],
       [0.63926236, 0.        , 0.1683962 , 0.13738393, 0.11353205,
        0.44550937, 0.1144353 , 0.11203857, 0.0977636 ],
       [0.08894391, 0.1683962 , 0.        , 0.01724109, 0.36172414,
        0.02186478, 0.22353113, 0.01319444, 0.01294449],
       [0.70218011, 0.13738393, 0.01724109, 0.        , 0.06598438,
        0.09843108, 0.06941839, 0.10582383, 0.07374102],
       [0.35886454, 0.11353205, 0.36172414, 0.06598438, 0.        ,
        0.08378407, 0.49501981, 0.10780812, 0.07617536],
       [0.46679106, 0.44550937, 0.02186478, 0.09843108, 0.08378407,
        0.        , 0.08559953, 0.07299823, 0.10957597],
       [0.33813901, 0.1144353 , 0.22353113, 0.06941839, 0.49501981,
        0.08559953, 0.        , 0.06822031, 0.08640308],
       [0.36950633, 0.11203857, 0.01319444, 0.10582383, 0.10780812,
        0.07299823, 0.06822031, 0.        , 0.06747319],
       [0.34050716, 0.0977636 , 0.01294449, 0.07374102, 0.07617536,
        0.10957597, 0.08640308, 0.06747319, 0.        ]])
Coordinates:
  * x        (x) <U8 'v_choice' 'j_choice' 'd_gene' ... 'vd_ins' 'dj_ins'
  * y        (y) <U8 'v_choice' 'j_choice' 'd_gene' ... 'vd_ins' 'dj_ins'

mdl_hb.plot_mutual_information(da_mi_scenarios)

<AxesSubplot:>

mdl_hb.plot_mutual_information(da_mi)

<AxesSubplot:>

Mean and Variance of values in scenarios dataframe

The simplest way to get the realization values of events in scenario is using the function IgorModel.get_realization_value_from_df_scenarios(df_scenarios, ‘event_nickname’)

help(mdl_hb.get_realization_value_from_df_scenarios)

Help on method get_realization_value_from_df_scenarios in module pygor3.IgorIO:

get_realization_value_from_df_scenarios(df_scenarios, event_nickname) method of pygor3.IgorIO.IgorModel instance

mdl_hb.get_realization_value_from_df_scenarios(df_scenarios_many, 'd_gene')

seq_index
998        GGGACAGGGGGC
998        GGGACAGGGGGC
998        GGGACAGGGGGC
998        GGGACAGGGGGC
998        GGGACAGGGGGC
             ...
703    GGGACTAGCGGGGGGG
703    GGGACTAGCGGGGGGG
703    GGGACTAGCGGGGGGG
703        GGGACAGGGGGC
703        GGGACAGGGGGC
Length: 10000, dtype: object

mdl_hb.get_realization_value_from_df_scenarios(df_scenarios_many, 'j_5_del')

seq_index
998    15
998    15
998    15
998    15
998    15
       ..
703     4
703     5
703     4
703     4
703     4
Length: 10000, dtype: int64

If the values are assigned to a column in scenarios dataframe (df_scenarios_many) the weighted average can be calculated using the IgorModel.w_mean_df_scenarios

ps_j_5_del = mdl_hb.get_realization_value_from_df_scenarios(df_scenarios_many, 'j_5_del')
df_scenarios_many['j_5_del_value'] = ps_j_5_del
# ps_j_5_del

help(mdl_hb.w_mean_df_scenarios)
help(mdl_hb.w_variance_df_scenarios)

Help on method w_mean_df_scenarios in module pygor3.IgorIO:

w_mean_df_scenarios(column_name: str, df_scenarios: pandas.core.frame.DataFrame) method of pygor3.IgorIO.IgorModel instance
    Return weighted mean with the normalized probabilities for each
    scenario (norm_scenario_proba_cond_seq)
    :param column_name: column name of df_scenario to calculate the average
    :param df_scenarios: Scenarios with normalize probability. Loaded with self.get_dataframe_scenarios()

Help on method w_variance_df_scenarios in module pygor3.IgorIO:

w_variance_df_scenarios(colname_1: str, df_scenarios: pandas.core.frame.DataFrame) method of pygor3.IgorIO.IgorModel instance
    Return weighted covariance with the normalized probabilities of the column names given for each
    scenario (norm_scenario_proba_cond_seq)
    :param colname_1: column name of df_scenario to calculate the weighted covariance
    :param colname_2: column name of df_scenario to calculate the weighted covariance
    :param df_scenarios: Scenarios with normalize probability. Loaded with self.get_dataframe_scenarios()

mdl_hb.w_mean_df_scenarios('j_5_del_value', df_scenarios_many)

4.157669017539739

mdl_hb.w_variance_df_scenarios('j_5_del_value', df_scenarios_many)

10.537416194979743

help(mdl_hb.get_observable_from_df_scenarios)

Help on method get_observable_from_df_scenarios in module pygor3.IgorIO:

get_observable_from_df_scenarios(observable_function, df_scenarios: pandas.core.frame.DataFrame) method of pygor3.IgorIO.IgorModel instance
    Return a pandas series with the calculated observable over the df_scenarios dataframe.
    :param observable_function: This function should use the varibles with self.realization
    :param df_scenarios: Scenarios dataframe loaded with self.get_dataframe_scenarios.

def f_total_insertions(ps_scenario):
    my_internal = np.zeros(len(mdl.parms['vd_ins']))
    vd_ins = mdl.realization(ps_scenario, 'vd_ins')
    dj_ins = mdl.realization(ps_scenario, 'dj_ins')
    # my_internal[vd_ins.id] = 1
    # return my_internal
    return vd_ins.value + dj_ins.value

def f_total_deletions(ps_scenario):
    v_3_del = mdl.realization(ps_scenario, 'v_3_del')
    d_5_del = mdl.realization(ps_scenario, 'd_5_del')
    d_3_del = mdl.realization(ps_scenario, 'd_3_del')
    j_5_del = mdl.realization(ps_scenario, 'j_5_del')
    return v_3_del.value + d_5_del.value + d_3_del.value + j_5_del.value

mean_insertions = mdl.w_average_function_df_scenarios(f_total_insertions, df_scenarios_many)
mean_deletions = mdl.w_average_function_df_scenarios(f_total_deletions, df_scenarios_many)

mean_insertions, mean_deletions

(10.202437426325016, 15.645906964009958)

def f_give_thename(ps_scenario):
    v_choice = mdl_hb.realization(ps_scenario, 'v_choice')
    return v_choice

mdl_hb.get_observable_from_df_scenarios(f_give_thename, df_scenarios=df_scenarios_many)

seq_index
998    TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998    TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998    TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998    TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998    TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
                             ...
703    TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703    TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703    TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703    TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703    TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
Length: 10000, dtype: object

df_scenarios_many['v_call'] = mdl_hb.get_observable_from_df_scenarios(f_give_thename, df_scenarios=df_scenarios_many)
df_scenarios_many

	scenario_rank	scenario_proba_cond_seq	v_choice	j_choice	d_gene	v_3_del	d_5_del	j_5_del	d_3_del	vd_ins	vd_dinucl	dj_ins	dj_dinucl	Mismatches	norm_scenario_proba_cond_seq	j_5_del_value	v_call
seq_index
998	1	0.783497	77	4	0	9	6	19	9	4	[1, 1, 2, 3]	7	[2, 2, 2, 2, 0, 2, 0]	[]	0.000801	15	TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998	2	0.076884	77	4	0	9	6	19	10	4	[1, 1, 2, 3]	8	[2, 2, 2, 2, 0, 2, 0, 2]	[]	0.000079	15	TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998	3	0.043634	77	4	0	9	7	19	9	5	[1, 1, 2, 3, 2]	7	[2, 2, 2, 2, 0, 2, 0]	[]	0.000045	15	TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998	4	0.023517	77	4	0	10	6	19	9	5	[1, 1, 1, 2, 3]	7	[2, 2, 2, 2, 0, 2, 0]	[]	0.000024	15	TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
998	5	0.021809	77	4	0	9	8	19	9	6	[1, 1, 2, 3, 2, 0]	7	[2, 2, 2, 2, 0, 2, 0]	[]	0.000022	15	TRBV7-4*01;GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGT...
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
703	6	0.021822	45	12	1	6	11	8	7	6	[2, 2, 1, 0, 3, 3]	5	[2, 0, 3, 1, 1]	[]	0.000030	4	TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703	7	0.016150	45	12	1	5	11	9	7	5	[2, 1, 0, 3, 3]	6	[2, 2, 0, 3, 1, 1]	[]	0.000022	5	TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703	8	0.013252	45	12	1	5	11	8	8	5	[2, 1, 0, 3, 3]	6	[2, 0, 3, 1, 1, 2]	[]	0.000018	4	TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703	9	0.013172	45	12	0	6	11	8	4	8	[2, 2, 1, 0, 3, 3, 2, 1]	4	[2, 0, 3, 1]	[]	0.000018	4	TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...
703	10	0.010846	45	12	0	5	12	8	4	8	[2, 1, 0, 3, 3, 2, 1, 2]	4	[2, 0, 3, 1]	[]	0.000015	4	TRBV4-1*01;ACTGAAGTTACCCAGACACCAAAACACCTGGTCAT...

10000 rows × 17 columns

ps_scenario = df_scenarios_many.iloc[0]
realiz = mdl_hb.realization(ps_scenario, 'v_choice')
realiz.value, realiz.id, realiz.name

('GGTGCTGGAGTCTCCCAGTCCCCAAGGTACAAAGTCGCAAAGAGGGGACGGGATGTAGCTCTCAGGTGTGATTCAATTTCGGGTCATGTAACCCTTTATTGGTACCGACAGACCCTGGGGCAGGGCTCAGAGGTTCTGACTTACTCCCAGAGTGATGCTCAACGAGACAAATCAGGGCGGCCCAGTGGTCGGTTCTCTGCAGAGAGGCCTGAGAGATCCGTCTCCACTCTGAAGATCCAGCGCACAGAGCAGGGGGACTCAGCTGTGTATCTCTGTGCCAGCAGCTTAGC',
 77,
 'TRBV7-4*01')