dnbaker / bmftools Goto Github PK

BMFtools (Barcoded Molecular Families tools) is a suite of tools for barcoded reads which takes advantage of PCR redundancy for error reduction/elimination. The core functionality consists of molecular demultiplexing at fastq stage, producing a unique observation for each sequenced founded template molecule. Accessory tools provide postprocessing, filtering, quality control, and summary statistics.

Requirements: gcc4.9+ samtools 1.2+

git clone https://github.com/ARUP-NGS/BMFtools --recursive
cd BMFtools
make

bmftools

bmftools <--help/-h>

bmftools <subcommand> <-h>

These tools are divided into four categories:

1. Core functionality
2. Manipulation
3. Analysis

####bmftools collapse bmftools collapse combines reads sharing barcodes into single observations respectively.

First, the barcodes are added to the comment fields of the fastqs and split the records into subsets based on the first characters in the barcode. Then, reads with exactly-matching barcode are collapsed, with a meta-analysis performed on each base call.

bmftools collapse inline collapses templates where both strands were sequenced, whereas collapse secondary lacks strand information.

####bmftools rsq bmftools rsq uses positional information to collapse reads sharing alignment signatures with close barcodes under the assumption that they came from the same original founding molecule but with errors in reading the barcode.

####bmftools cap Caps quality scores using barcode metadata to facilitate working with barcode-agnostic tools.

####bmftools filter Filters or splits a bam file based on a set of filters. These can be inverted with -v (analogous to grep).

Filters:

• Fail reads with insufficient mapping quality.

• Fail reads with insufficient family size.

• Fail read pairs by aligned fraction.

• Fail reads outside of a bed region.

• Fail reads without all bits in given parameter in the sam flag field.

• Fail reads with any bits in given parameter in the sam flag field.

####bmftools vet Curates SNV calls from a tumor/normal variant call file using barcode metadata from the bams used to produce the variant call file.

####bmftools depth Calculates depth of coverage across a bed file using barcode metadata.

####bmftools target Calculates on-target fraction for bed file using barcode metadata.

####bmftools err Calculates error rates by a variety of parameters. Additionally, pre-computes the quality score recalibration for the optional collapse recalibration step.

####bmftools famstats Calculates summary statistics related to family size and demultiplexing.

####bmftools stack A maximally-permissive variant caller using molecular barcode metadata analogous to samtools mpileup.