Predict structural variation from mate-pair mapping

[jeffreybarrick]

Implement in C++ and use R for some visualization.

• [ ] Provide command line input functionality for linking together FASTQ files as containing paired reads. Probably connect them via an unusual character combination that is OK in the shell and then split the argument, something like R1.fastq%R2.fastq
• [ ] Add bowtie2 commands allowing mate-pair mapping. Be sure this does not interfere with split-read JC evidence steps.
• [ ] Estimate the normal distribution of insert lengths and orientations from the data and generate QC graphs and output files.
• [ ] Find consistent outliers and define as new MP evidence type.
• [ ] Integrate MP evidence with predicting mutations: for example IS element insertions, deletions, and amplifications.

[jeffreybarrick]

Here are a couple of LTEE samples mapped in paired-end mode to use for testing different approaches: https://barricklab.org/release/tmp/paired-end-mapping.tar