cdpybio 0.0.3

A collection of useful methods for working with various bioinformatics data, software output files, etc.

cdpybio

Python module with various submodules tailored to specific tools, analyses, etc. that I perform across projects. Documentation below is only an introduction, most functions are documented with docstrings.

Dependencies

• Biopython

• HTSeq

• numpy

• pandas

• pybedtools

• pysam (v0.6 or greater)

• PyVCF

• py.test for testing

Submodules

bedtools

Methods for working with bed files. I make heavy use of pybedtools.

cghub

This submodule contains some useful methods and classes for dealing with data through CGHub. I’ll go through the classes and methods.

GTFuseBam

A GTFuseBam object is a single bam file from CGHub mounted with GTFuse. You can mount and unmount the bam file as you’d like.

ReadsFromIntervalsBam

This object takes a GTFuseBam object and a set of intervals and obtains the reads from those intervals in the CGHub bam file and writes them to a local bam file.

• reads_from_intervals: Obtain the reads from the given intervals.

ReadsFromIntervalsEngine

This is an engine that runs in the background and obtains reads from intervals for a given set of samples. The main process that runs the engine shares the thread with your python session but I use the multiprocessing module to farm out different bam files to different threads so you can obtain reads from multiple bam files simultaneously. This class can be extended.

FLCVariantCallingEngine(ReadsFromIntervalsEngine)

This engine extends the ReadsFromIntervalsEngine and performs variant calling after obtaining the reads. Currently implemented to work in the Frazer lab computing environment although it would be easy to change for a different computing environment.

TumorNormalVariantCall

Class that wraps the results of a variant calling job for a tumor/normal pair.

express

This submodule has some methods that are useful for dealing with the output from the RNA-seq expression estimation tool eXpress.

• combine_express_output: Combine multiple eXpress output files into a single pandas dataframe. You can choose which column to combine. You can also aggregate values by gene ID (eXpress estimates transcript expression) if you provide a mapping from transcript IDs to gene IDs. See gencode.make_transcript_gene_se.

gencode

Functions for parsing the Gencode gene annotation into various files that are easier to work with.

• make_transcript_gene_se: Make a file with a simple mapping from transcript IDs to gene IDs.

• make_gene_info_df: Make a file indexed by gene ID that has some simple information about each gene.

• make_splice_junction_df: Make a file indexed by splice site that has information about each splice site such as gene, strand, acceptor, donor, etc.

general

Some methods that are generally useful.

mutect

Methods for working with MuTect output.

pysamext

Provides extended functionality on top of pysam.

star

• def read_sj_out_tab: Read sj.out.tab file from STAR and parse it.

• def read_external_annotation: Read file with junctions from some database. This does not have to be the same splice junction database used with STAR. The file must have some specific columns (see docstring). Compatible with output from the gencode.make_splice_junction_df.

• def combine_sj_out: Combine SJ.out.tab files from STAR by filtering based on coverage and comparing to an external annotation to discover novel junctions.

• def make_logs_df: Make pandas DataFrame from multiple STAR Log.final.out files.

variants

Useful for tools for working with DNA variants.