Bam processing (WIP)

[averagehat]

Blocked by VDBWRAIR/bioframes#5, see also #7 (comment)
also need some kind of simple test that ensure that jip is co-operating correctly, ala this but more straightforward I reckon

I added a dependency on bioframes, I will add a dependency to biotest, will probably want to contribute a way to generate fasta records and maybe even something for VCFs there.

This includes:

• [ ]tagging & sorting & indexing bam with a modified version of ngs_mapper's tagreads.py (which writes to a new file/stdout and tags all reads as required by freebayes)
• jip job for freebayes
• new implmementation of consensus.py
• tests for jip jobs
• tests for consensus.py

[averagehat]

NB: need to fix the python contracts, forgot how to get it to work with seqrecord and non-list iterables

[averagehat]

A consensus can have deletions at either the end of the reference. When all of the reads support this, none of that position gets reported in the pileup, and freebayes won't catch that the consensus is shorter at the ends (at least I think this is what happens).
The way to get around this is to trim the consensus by those ends of the pileup which are empty.

[averagehat]

Freebayes entries can look like this:

H3N2/CY074920/Managua/2010/PA_3 207 .   AC  GT,AT   8859.33 .   AB=0.432024,0.564955;ABP=16.295,15.1404;AC=1,1;AF=0.5,0.5;AN=2;AO=143,187;CIGAR=2X,1M1X;DP=331;DPB=331;DPRA=0,0;EPP=39.4698,4.41537;EPPR=0;GTI=0;LEN=2,1;MEANALT=3,3;MQM=60,60;MQMR=0;NS=1;NUMALT=2;ODDS=716.828;PAIRED=0.503497,0.754011;PAIREDR=0;PAO=0,0;PQA=0,0;PQR=0;PRO=0;QA=4862,6499;QR=0;RO=0;RPL=110,89;RPP=93.0429,3.95088;RPPR=0;RPR=33,98;RUN=1,1;SAF=101,129;SAP=55.8697,61.5472;SAR=42,58;SRF=0;SRP=0;SRR=0;TYPE=mnp,snp;technology.ILLUMINA=1,1 GT:DP:RO:QR:AO:QA:GL    1/2:331:0:0:143,187:4862,6499:-922.526,-528.256,-485.208,-394.293,0,-338.001

Note that although base-by-base information is not provided, information is provided for both possible mutations, which is good.

[averagehat]

I propose testing this by adding it to the map pipeline in #20
ready to merge

[averagehat]

I need to think about whether the example here undercuts this implementation. I don't think it does.

[averagehat]

Notice that the build errors start after a change to the readme file in 272a115. This is another setuptools issue. There hasn't been an update to the pyvcf package in 3 months. Also, the distribute package has been deprecated for some time? I don't know. @necrolyte2 have you seen this one before?

[necrolyte2]

I've already forgotten what we did last time that fixed that type of error.
I hate setuptools/distribute and just want to punch it in the face