<div style="width:680px;margin:0 auto;text-align:justify;" >

  <div style="margin-top: 50px;">
    <b> THIS IS STILL THE MITOS(1) DOCUMENTATION.</b> 
  </div>


  <div style="margin-top: 50px;">
    Welcome to <b>MITOS</b>, a free web server for the annotation of metazoan
    mitochondrial genomes. Using <b>MITOS</b> has a number of advantages
    compared to manual curation:
  </div>

  <div style="width: 465px; margin-left: auto; margin-right: auto; text-align: left; margin-top: 10px;" >
    <ul>
      <li> <b>reliable and consistent analysis</b> of metazoan mitochondrial genomes</li>
      <li> <b>simple usage</b> - selection of cutoff values not necessary</li>
      <li> <b>advanced usage</b> - for users who are not satisfied with the default values</li>
      <li> <b>widely used file formats</b> for sequence upload and download of the results</li>
      <li> <b>state-of-the-art methods</b> - novel structure-based covariance models for ncRNA annotation</li>
      <li> <b>unmatched sensitivity</b> in tRNA prediction and <b>improved detection</b>
     	of rRNA 5' and 3' ends</li>
    </ul>
  </div>

  <br/>
  Click <a href="http://{mitosurl}">here</a> to start using <b>MITOS</b> or
  continue reading the following sections to learn more about <b>MITOS</b> and how to use it:

  <div style="width: 420px; margin-left: auto; margin-right: auto; text-align: justify; margin-top: 10px;">

    <a href="#overview">Overview</a><br/>
    <a href="#input">Input</a><br/>
    <a href="#output">Output</a><br/>
    <a href="#faq">FAQ</a><br/>
    <a href="#example">Example</a><br/>
    <a href="#advanced">Advanced Usage</a><br/>
    <a href="#uses">Studies using <b>MITOS</b></a><br/>
    <a href="#privacy">Privacy</a><br/>

  </div>

  <br/>
  <h2 id = "overview">Overview</h2>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
    <b>MITOS</b> is a web server for the automatic annotation of metazoan
    mitochondrial genomes. <b>MITOS</b> allows a reliable and consistent
    annotation of proteins and non-coding RNAs. The analysis steps are as
    follows:<br/>
    <div style="margin-top: 10px; margin-left: 30px; margin-right: 50px;" >
      <ul>
	<li> Candidate <i>protein coding genes</i> are found by detecting congruences in the
	  results of blastx searches against the amino acid sequences of the annotated
  	  proteins of metazoan mitochondrial genomes found in the NCBI RefSeq 39. A
  	  postprocessing step detects start and stop codons, duplicates, and hits belonging to
  	  the same transcript, e.g. frame shift or splicing.<br/><br/>
	</li>
	<li> <i>tRNAs</i> are annotated using MITFI, i.e. novel structure-based covariance models
      as described in <a href="http://dx.doi.org/10.1093/nar/gkr1131" target="_blank">J&uuml;hling, et al. Nucleic Acids Research, 2012, 40(7):2833-2845</a>.
      This approach was shown to have an unmatched sensitivity (outperforming ARWEN and tRNAscan-SE, respectively)
  	  and a precision higher than ARWEN and equivalent to tRNAscan-SE.
  	  <br/><br/>
	</li>
	<li> <i>rRNA</i> annotation is performed using structure-based covariance
#   	  models that have been developed similarly to the tRNA models.
  	  Structural considerations improve 5' and 3' end predictions of the rRNAs.<br/><br/>
	</li>
	<li>In a final step, conflicts are resolved and the outcome is prepared for
  	  visualization.
	</li>
      </ul>
    </div>
  </div>

  <h2 id = "input">Input</h2>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
    Annotating mitochondrial genomes with <b>MITOS</b> is rather simple and
    straightforward, unburdening the user from having to select cutoff values
    and the like. All that must be provided is:
    <ol>
	<li>
	  a <a href="https://en.wikipedia.org/wiki/FASTA_format" target="_blank">FASTA</a> file,
	  containing the genome to be annotated and<br/><br/>
	</li>
	<li>
	  the appropriate  <a href="http://www.ncbi.nlm.nih.gov/Taxonomy/Utils/wprintgc.cgi" target="_blank">genetic code</a>.
	</li>
      </ol>
    <br/>
    The user does not have to provide a name or e-mail address. However, if one
    chooses to do so, a link to the results page will be sent to the given
    address upon completion of the job.
    In either case, after submission of the job the user will be directed to a
    page displaying the status of the job, a link that allows the deletion of the job, and
    as soon as the job is finished the results will be shown on the same page.
  </div>

  <h2 id = "output">Output</h2>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
    Results are provided in commonly used file formats:
    <a href="http://genome.ucsc.edu/FAQ/FAQformat.html#format1" target="_blank">BED</a>,
    <a href="http://genome.ucsc.edu/FAQ/FAQformat.html#format3" target="_blank">GFF</a>,
    <a href="https://en.wikipedia.org/wiki/FASTA_format" target="_blank">FASTA</a>, and
    <a href="http://www.ncbi.nlm.nih.gov/genbank/tbl2asn2#tbl" target="_blank">TBL</a> format and
     a visual overview of the results is presented.  Additionally, a file
     containing the gene order, i.e. the gene names in the order of their
     appearance on the genome, that can be used for genome rearrangement
     analyses (e.g. with <a href="http://pacosy.informatik.uni-leipzig.de/crex/" target="_blank">CREx</a>) is
     provided. Furthermore, the raw data, i.e. all files that have been generated by <b>MITOS</b>,
     is available as zip archive.
    <br/>
  </div>

  <div style="margin-top: 0px; margin-left: 20px; margin-right: 50px;" >
  <h3 id = "scores">Evaluating the Output</h3>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
    <b>MITOS</b> calculates <i>e-values</i> for ncRNA and <i>quality values</i> for protein coding
    gene predictions. <a href="mito/data/stat-pshs-logx-h.pdf">Here</a>, we provide plots with the
    quality value distributions for
      <ul>
	<li> the initial hits found for the 13 protein coding gene families in RefSeq 39 (<i>norm</i>),
	</li>
	<li> the sequences without the respective gene (<i>cut</i>), and
	</li>
	<li> the trinucleotide frequency preserving permutations (<i>shuffle</i>).
	</li>
      </ul>
      Within the plots the best 939 (50% the size of RefSeq 39; dashed line) and 1878 (solid line)
      <i>quality values</i> of the initial hits are indicated.
      These plots might be used for estimating the relative quality of the results.
    </div>
  </div>


  <h2 id = "faq">FAQ</h2>
  <div style="text-align: left; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
    <b>Q:</b> <i>How long does it take for <b>MITOS</b> to annotate my mitochondrial genome?</i><br/>
    <b>A:</b> MITOS needs about 1.5h for a single mitogenome of average length.
	If multiple users submit jobs they are put in a queue. The jobs in the queue are
    processed in the order of their arrival and two jobs can be processed in parallel.
    So, if your job is queued the time until the job will be started might be roughly as:
    queue position ~ hours to wait.
    <br/>
    <br/>
    <b>Q:</b> <i>How can I share the results of <b>MITOS</b> with colleagues?</i><br/>
    <b>A:</b> By sending them the URL shown in the browser when you access your results.
    <br/>
    <br/>

    <b>Q:</b> <i>How secure is the data that is uploaded and the results? Who can access it?</i><br/>
    <b>A:</b> Your data is only accessible by persons knowing the precise URL. This is
    initially only the person who submitted the job. Thus this person can control who may
    access the data and results by keeping the URL a secret or distributing it.
    The uniqueness of the URL is ensured by a unique job identifier (which is generated using the
    SHA-256 method from the submitted data). It is extremely unlikely (virtually impossible)
    to find an identifier at random. Thus, a high security level is ensured and
    at the same time it is still easy for you to distribute the results to colleagues.
    <br/>
    <br/>

    <b>Q:</b> <i>How long are the results kept? </i><br/>
    <b>A:</b> The results are deleted automatically 1 month after the submission, but they can be
    regenerated by restarting the job which is possible via the result link (for jobs created after
    April 2013).
    <br/>
    <br/>

    <b>Q:</b> <i>Do you provide results in other formats?</i><br/>
    <b>A:</b> Currently not. However, if you think we should offer an additional
    format please send us a message: <script type='text/javascript'> var pref = '&#109;a' + 'i&#108;' + '&#116;o'; var attribut = 'hr' + 'ef' + '='; var first = '%6D%69%74%6F%73'; var at = '%40'; var last = '&#x62;&#x69;&#x6F;&#x69;&#x6E;&#x66;&#x2E;&#x75;&#x6E;&#x69;&#x2D;&#x6C;&#x65;&#x69;&#x70;&#x7A;&#x69;&#x67;&#x2E;&#x64;&#x65;'; var first2 = '&#x6D;&#x69;&#x74;&#x6F;&#x73;'; var at2 = '&#x40;'; var last2 = '&#98;&#105;&#111;&#105;&#110;&#102;&#46;&#117;&#110;&#105;&#45;&#108;&#101;&#105;&#112;&#122;&#105;&#103;&#46;&#100;&#101;'; document.write( '<a ' + attribut + '\'' + pref + ':' + first + at + last + '\'>' ); document.write( first2 + at2 + last2 ); document.write( '<\/a>' ); </script> <noscript> <div style='display:none; '>are-</div><div style='display:inline; '>&#x6D;&#x69;&#x74;&#x6F;&#x73;</div><div style='display:none; '>-xya34</div><div style='display:inline; '>[at]</div><div style='display:none; '>ddks-</div><div style='display:inline; '>&#98;&#105;&#111;&#105;&#110;&#102;&#46;&#117;&#110;&#105;&#45;&#108;&#101;&#105;&#112;&#122;&#105;&#103;&#46;&#100;&#101;</div> </noscript>.
    <br/>
    <br/>

    <b>Q:</b> <i>I don't have a complete genome...</i><br/>
    <b>A:</b> Basically, no problem. <b>MITOS</b> can also annotate fragments. But you should be
    aware of MITOS' strategy which tries to enforce a copy for each of the default metazoan mitogenes.
    This might evoke low scoring spurious hits.
    <br/>
    <br/>

    <b>Q:</b> <i>How do I cite <b>MITOS</b>?</i><br/>
    <b>A:</b> Please cite:<br/>
	<i>M. Bernt, A. Donath, F. J&uuml;hling, F. Externbrink, C. Florentz, G. Fritzsch, J. P&uuml;tz, M. Middendorf, P. F. Stadler</i><br>
	<b>MITOS: Improved de novo Metazoan Mitochondrial Genome Annotation</b><br>
	Molecular Phylogenetics and Evolution 2013, 69(2):313-319
	<a href="http://dx.doi.org/10.1016/j.ympev.2012.08.023" target="_blank">link</a>
    <br/>
    <br/>
    <b>Q:</b> <i>What can I do if <b>MITOS</b> produced an error or is behaving unexpectedly? </i><br/>
    <b>A:</b> It would be nice if you inform us by sending an email to {mail}.
    <br/>
    <br/>
  </div>

  <h2 id = "example">Example</h2>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">

    This example is based on the complete mitochondrial genome of
    <i>Arborophila rufipectus</i> (NC_012453). You can download
    the sequence in FASTA format <a href="NC_012453.fasta">here</a>.
    Precomputed results from <b>MITOS</b> for this example can be found
    <a href="result.py?hash=example">here</a>. If you
    want to compare it with the annotation found in RefSeq,
    <a href="NC_012453.gff2">here</a> is a GFF compiled
    from the original GenBank file (<a href="http://www.ncbi.nlm.nih.gov/nuccore/NC_012453.1" target="_blank">NC_012453.1</a>).<br/> <br/>

    <h4 id = "upload">UPLOAD</h4>
    <div class="helptxt">
      All you have to provide is the sequence in FASTA (Pearson) format and
      the selection of the correct genetic code. The job is submitted
      by clicking the <i>Proceed</i> button.
      If you enter a valid email address you will receive a notification as
      soon as the analysis is complete.
      The notification email will contain a link to the <b>MITOS</b> results
      of the job and mention the job identifier (or filename if not given).
      The job identifier will also be given on the job settings page.
      Optionally you can provide a name wich will be used to adress you.
      <br/>
      If all of our resources are currently in use, your job will be put in a queue
      and you are notified about your position. Here you also have the possibility to
      delete your job.<br/>
      <br/>
      Once your job is being processed you will be notified about this.
      Cancelling the job is not possible as soon as it is executed.<br/>
      The page will reload every 30 seconds until the results are available. If
      you have provided a valid email address you can close this window.
      Otherwise you have to bookmark it or leave it open until the analysis is
      complete.<br/>
    </div>

    <img src="mito/input.png" alt="Input mask" class="helpimg">
    <img src="mito/analysis.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

    <h4 id = "results">RESULTS</h4>

    <div class="helptxt">
    Once the annotation is ready, you will be redirected the results page. This page
    has a menu on the left that offering several downloads and the results in tabular and
    graphical form on the right.
    </div>

	<div style="clear: both;"> </div>


    <div class="helptxt">
    The central part is the results table where you find all genes predicted by <b>MITOS</b> and their
    position in the provided sequence. The coordinates are as defined for the
    GFF format, i.e. the first base of the sequence is numbered 1 and start and stop of the feature are included.
    For tRNAs and rRNAs we provide secondary structure plots in svg and postscript format.
    </div>

    <img src="mito/resulttop.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

    <div class="helptxt">
      Below the table you will see a visualization of the annotation.
      Genes located on the plus strand are drawn in the upper part. Genes annotated on the
      minus strand are shown in the lower region. A small vertical line is
      drawn every 1,000 nt. A click on the image will present a larger
      annotated version.<br/>
      Note that the legend of the plot indicates which feature types have been searched
      by <b>MITOS</b> (tRNA, rRNA, protein). In our example all three types have been searched - which is the default setting of <b>MITOS</b>.

      On the very bottom of the page <b>MITOS</b> reports peculiarities and problems of the
      results which may need your attention. Most importantly genes that could
      not be located are listed. This may happen if there is no signal for the gene or if
      its predicted position conflicts with other genes (e.g. if a gene is annotated to long).
      Furthermore cases of genes that have been found as copies or in multiple parts are listed.
      For both cases we provide detailed data that support the analysis of these cases
      (described below).
    </div>

    <img src="mito/resultimg.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

    <div class="helptxt">
    In the upper left you will find the <b>Downloads</b> section which contains
    links to different file formats to download the annotation of your sequence.
    Below, in the <b>Raw Data</b> section a zip archive containing
    all raw data that has been created during the annotation process,
    i.e. unfiltered BLAST outputs, results from the HMM search etc.
    Furthermore, the protein and ncRNA plots provide an easy overview on the raw data,
    i.e. the quality values of the initial hits from <b>MITOS'</b> protein prediction
    method and the e-values of the original infernal predictions (details below).
    <br/>
    </div>

    <img src="mito/resultfiles.png" alt="analysis" class="helpimg">

	<div style="clear: both;"> </div>

    <div class="helptxt">
    <b>MITOS</b> annotates two copies of <i>nad6</i> for this sequence:
    the first copy on the plus strand (736-1,155) and the other on the minus
    strand (16,141-16,659). You will also notice that the
    second copy is roughly 100 nt longer than the other one. Now we
    need to take a look at one of the result files to confirm whether we have a
    real duplication event.
    </div>

    <img src="mito/resultbottom.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

    <div class="helptxt">
    Download, for example, the BED file and open it
    with an editor of your choice. (Keep in mind that coordinates in
    BED files start with 0 and the stop is <b>not</b> included.)

    <div style="margin-top: 10px;"></div>
    <b>MITOS</b> annotates protein coding genes by detecting congruences
    in the results of BLASTX searches. That is, in short, the <i>e-values</i>
    of BLASTX searches aiming at the same target are summed up position-wise
    (see also: <a href="#scores">Evaluating the Output</a>). Regions
    defined by stretches of positions with a "good" score are considered as
    connected and part of the respective protein coding region. This
    enables <b>MITOS</b> to detect frame-shifts and, by applying a
    greedy method, duplication events. However, in rare cases, this
    greedy procedure annotates an unlikely copy.<br/> This can easily
    be identified by evaluating the scores given in the annotations
    (or the protein plots).
    Certainly, <i>e-values</i> (and therefore <i>quality values</i>) of two
    different BLAST searches cannot be compared. Nevertheless, they
    give us a good indication whether we have a true copy or just a
    region with a slight similarity.<br/>
	As we can see, the putative <i>nad6</i> duplicate on
    the minus strand has a much higher score than the one on the plus
    strand. Additionally, most other protein coding regions exhibit
    similar high quality values. From this we can conclude that the
    region on the minus strand contains a true <i>nad6</i> gene, but the
    other one is unlikely and needs further examination.
    </div>

    <img src="mito/nad6a.png" alt="analysis" class="helpimg">

    <img src="mito/nad6b.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

    <div class="helptxt">
	Further possibilities for a more detailed analysis of the results is given
	by the protein and ncRNA plots that are linked in the "Raw data" section.
	The protein plot shows for each gene and each position the quality value
	if it is above the threshold. Different colors are used to differentiate
	the genes. Basically, the initial hits used in MITOS correspond to the
	"mountains" in this plot. Thus this plot visualizes the signal from the
	BLAST searches. The arrows shown on the top of the plot show the annotation
	reported by <b>MITOS</b>. Note that the quality values are shown on a log scale.
	In the protein plot for the example one can clearly see the inferior quality values
	of the spurious nad6 hit.
	<br>
	The ncRNA plot shows the hits reported by infernal where conflicts among
	tRNAs (respectively rRNAs) are already removed. The plot differentiates between
	the hits from the global and glocal (if present) search by line type.
	Furthermore the hits are separated in priorities (1 or 2). Priority "1" refers
	to features set in the first round when MITOS takes the prediction hit of each
	feature, and priority "2" refers to all other features set afterwards in the
	remaining unassigned regions.
	Note the reverse log scale for the e-value.
    </div>

	<img src="mito/protplot.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

    <div class="helptxt">
	In order to retrieve the parameters used for the annotation of your job
	we provide the <i>job settings</i> page. The link is the last item in the menu
	on the results page. The setting are shown in tabular form. The heading shows
	job identifier that was provided.
    </div>

	<img src="mito/settings.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>

  </div>
  <h2 id = "advanced">Advanced Usage</h2>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
    The default values implemented in <b>MITOS</b> are tested
    thoroughly and work well for different taxonomic lineages. If,
    however, you are not satisfied with the result of the annotation
    or if you are only interested in certain gene types, <b>MITOS</b>
    offers an <i>advanced mode</i>. To change the annotation
    parameters, press the <i>Advanced</i> button below the input mask.

    <div style="clear: both;"> </div>
    <h4>GENERAL SEARCH SETTINGS</h4>
	<b>Multi Fasta</b><br/>
	Per default the <b>MITOS</b> web server accepts only files containing a single
	fasta formatted sequence. This may be changed with the Multifasta
	checkbox. Please use this option carefully, considering the following
	points:
	<ul>
	<li> If a multifasta file is uploaded for each sequence in that file a job
		will be created. Your web browser will be redirected to the results page
		of the job corresponding to the last sequence in your file.</li>
	<li> In order to be able to access the results of the other sequences you
		have to specify an email address. Note that, MITOS will send a result
		notification mail for each sequence (which might have spam like character).</li>
	<li> Please consider that MITOS currently needs about 1-2 hours for a
		typical mitochondrial genome. Hence a larger amount of jobs will create
		a considerable computational demand which might delay also jobs of other
		users. You can check the current workload of the MITOS web server on the
		<a href="http://{mitosurl}/stats.py">Status and Statistics page</a>.</li>
	<li> For larger data sets we could provide dedicated computational
		resources. Also more convenient access to larger sets of results can be
		organized.</li>
	<li> In case of questions you might contact us at {mail}.</li>
	</ul>
	<br/>

    <b>Feature types</b><br/>
    Here you can select if you want to annotate all three types of
    genes or only a subset. If you are only interested in protein
    coding genes, untick the boxes for tRNAs and rRNAs. This will
    also decrease the time needed for annotation dramatically.
    See the <b>MITOS</b> article for a further explanation of the parameters.<br/>

      <h4>PROTEIN SEARCH PARAMETERS</h4>
      <div class="helptxt">
      <b>BLAST E-value Exponent</b>
      Change this value to specify the statistical significance
      threshold for considering matches in the BLASTX search.
      The value entered here is the negation of the exponent of the E-value
      threshold that should be used by BLAST, i.e. a value X gives an E-value of 10^(-X).
      <br/>
      <br/>
      <b>Cutoff</b>
      Minimum allowed quality value (in percent) of the maximum
      quality value per reading frame. A higher values correspond to
      shorter protein prediction and therefore reduced risk for
      conflicts with other features <br/>
      <br/>
      <b>Maximum Overlap</b>
      Maximum allowed overlap (in percent) of the smaller
      feature.<br/>
      <br/>
      <b>Clipping Factor</b>
      Clipping is started if overlapping prediction of hits with the same
      name differ by less than a factor X in their quality value.<br/>
      <br/>
      <b>Fragment Overlap</b>
      Maximum fraction (of the shorter feature) allowed that two hits
      overlap in the query to be counted as fragments.<br/>
      <br/>
      <b>Fragment Quality Factor</b>
      Maximum factor by which fragments may differ in their quality
      scores. Higher values allow that parts of a gene can differ more
      in their quality.<br/>
      <br/>
      <b>Start/Stop Range</b>
      Range (number of amino acids) in which start and stop codon are
      searched.
      <h4>FINAL STEP PARAMETERS</h4>
      <b>Final Maximum Overlap</b>
      Maximum number of nucleotides by which genes of different
      types may overlap. Applies to merging of the final
      predictions.
    </div>

    <img src="mito/input_adv.png" alt="analysis" class="helpimg">

    <div style="clear: both;"> </div>
    </div>

  <h2>Raw Data</h2>
  <div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
	The raw data contains the following:
	<dl>
	<dt>blast/</dt><dd>The BLAST results. The directory "prot" contains the output files produced by BLAST in tabular format. The file "blast.dat" is a auxiliary file that is used to generate the protein plot. </dd>
	<dt>mitfi-global/ and mitfi-local/</dt><dd>The results of MiTFi (tRNAout.nc and rRNAout.nc) and infernal (.cmout file) for the global and glocal search search.</dd>
	<dt>plots/</dt><dd>All plots: protein plot, ncRNA plot, and all ncRNA secondary structures.</dd>
	<dt>result</dt><dd>File containing the results. Positions are 0-based and start&amp;stop position are included. Columns are: id, type, name, prediction-method, start, stop, strand, anticodon, copy-number, part-number, structure, anticodon-position.</dd>
	<dt>sequence.fas</dt><dd>The original sequence.</dd>
	</dl>
  </div>

  <h2>Original data used in <b>MITOS</b></h2>
<div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
To allow for reproducibility we publish the used original data.
      <ul>
	<li> The seed alignments and structural models for the ncRNA can be found <a href="mito/ncRNA.zip"> here</a></li>
	<li> The aminoacid sequences used to find the protein coding sequences are deposited <a href="mito/RefSeq39.zip">here</a>.</li>
      </ul>
</div


<h2><b>MITOS</b> source code</h2>
<div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">
The sources of MITOS are available on <a href="https://gitlab.com/Bernt/MITOS">gitlab</a>. 
</div

  <br/>
  <h2 id = "uses">Studies using <b>MITOS</b></h2>
<div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">

<ol reversed="reversed">

<!--
<li>.
.
 <a href="http://dx.doi.org/" target="_blank">link</a>.
</li>
-->

</ol>

</div>

<br>
<h2 id = "otheruses">Other studies citing <b>MITOS</b></h2>
<div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">

<ol reversed="reversed">

</ol>
</div>

<h2 id = "privacy">Privacy</h2>
<div style="text-align: justify; margin-top: 10px; margin-left: 20px; margin-bottom: 10px;">

Submitted sequence data is not shared with others.
Sequence data and results are deleted after one month.
<br><br>


The MITOS webserver collects usage data.
We do not share this data with others and we respect the <a href="http://de.wikipedia.org/wiki/Do_Not_Track">do not track</a> option of your browser.
Independent of this we give you the possibility to activate / deactivate tracking for this site:<br>

<iframe frameborder="no" width="600px" height="200px" src="http://piwikmitos.bioinf.uni-leipzig.de/index.php?module=CoreAdminHome&action=optOut&language=en"></iframe>
</div>
</div>
