## 0.1.5.5
## 0.1.5.4
## 0.1.5.3
## 0.1.5.2
1. Fix relative path bug

## 0.1.5.1
## 0.1.5
1. Now pair end reads in split bam file will be matched to each other. 
2. Fixed bug in clipped bam report and the final report.

## 0.1.4
1. Now move identifiers in reads' name to tags:
        MT: Mate information. 0: mate1; 1: mate2; 3: single end mapping. 
        FG: Fragment order.
        LC: Clipping point from the reads head.
        RC: CLipping point from the reads tail.

## 0.1.3
1. Now use sort in pysam to sort all bam files.
2. Now useless tags will be removed.

## 0.1.2
1. Add 0.05*read_length as mismatch threshold.
2. Now original output bam doesn't contain any unmapped reads.

## 0.1.1
1. Add -u to LiBis first round mapping.
2. Add reminder for each step.

## 0.1.0
1. Add fast mode when BSMAP allows -U to output unsorted reads.
2. MOABS>=1.3.8.5

## 0.0.15
1. Combined multiple diction in mapreduce
2. Reduce the complexity in unmapped reads selection.
3. Changed to order of unmapped reads in unmapped fastq. Now fragments from one unmapped are adjacent and are ranked by file order from 0 to MAX.

## 0.0.14
1. Add RELEASE to the pypi package

## 0.0.13
1. Remove 'sort' in mapreduce step

## 0.0.12
1. add -rc. 
2. All temp fastq files are in GZ format now.

## 0.0.9
1. add -module, -mm ,and -pnf.
2. All temp SAM files are replaced by pysam.

## 0.0.7
Now support installation by pip and conda

## 0.0.3
1. Fixed reads name bug, remove the redundant modification of reads name like "@SRR001666.1". Keep the removal of "/1" or "/2" at the end of the reads.
2. Add left cut length and right cut length to the reads name. Now the reads name contains 4 fields at the end divided by "_":
3. The first number represent the rank of fragment from the reads(For example, if there are two fragments clipped from one reads, the second field will be 0 and 1).
4. The second number means the mate, which file does the read come from.
5. The third number means the left cut length
6. The forth number means the right cut length

## 0.0.2
1. Decided the name of the software: LiBis, which stands for Low input Bisulfite sequencing
2. Add -mcall, -plot, -nc, -fullmode
3. normal mode of bsmap now use label as the filename for generated bam.

## 0.0.1
1. Remove all part fastq/bam/sam. Merge all fragments into one file to speed up the computation.
2. Extension requires the overlap between two fragments.
3. Add -bam option to allow users use their own bam file for the first stage mapping.



