output_fields.md

Bamstats Output Fields

This document describes the fields of the different sections of a Bamstats output file.

General

The general section contains mapping statistics collected from the input BAM file.

Fields

protocol

The protocol used for sequencing, extracted from the SAM flags. It can be either SingleEnd or PairedEnd.

reads

Statistics and counts for single reads. In case of PairedEnd data each mate is counted independently.

total

The total number of reads, in samtools view -c -F256. It corresponds to the sum of unmapped reads and all the values in the mapped reads object.

unmapped

The number of unmapped reads, as in samtools view -c -f4.

mapped

This is an object containing the number of mapped reads grouped by the number of hits each read has (NH tag in the SAM format). The sum of these values gives the total number of mapped reads.

mappings

An object containing the following information on the alignments:

• the global ratio of the total number of mappings over the total number of mapped reads, representing the average number of hits per read
• the total number of mappings as in samtools view -c, including muliple hits for each read

pairs

Statistics and counts for read pairs, additionally reported if the data proocol is PairedEnd. The same fields as in the reads section, except for the mappings object, are included but referring to read pairs.

Fields

insert_sizes

An object containing the count of mapped pairs grouped by the corresponding insert size length.

Genomic Coverage

The genomeCoverage section contains metrics for genomic coverage based on the provided annotation. The counts are computed for continuous and split reads. For split reads the aligned blocks are considered separately. An aggregated report with the total counts is also collected.

An additional genomic coverage section for uniquely mapped reads called genomeCoverageUniq is additionally reported in the output file when the --uniq (or -u) command line option is used.

Fields

exon

Reads mapping to an exonic region. Reads must be totally included. For split reads, all the blocks must be included in an exonic region.

intron

Reads mapping to an intronic region. Reads must be totally included. For split reads, all the blocks must be included in an intronic region.

exonic_intronic

Reads overlapping exon-intron junction. For split reads, any of the blocks can either overlap the junction, map to an exon or an intron.

intergenic

Reads mapping to an intergenic region. Reads must be totally included. For split reads, all the blocks must be included in an intergenic region.

others

Reads mapping to regions different from the ones described above. For split reads, this can also reported for unexpected regions combination of the alignment blocks.

RNAseq

The rnaseq sections contains metrics computed following the recommendations from the IHEC Assay Standards working group.

Fields

intergenic

The number of reads mapping to intergenic regions. Reads don't need to be totally included into the region and the whole read is used in case of split reads, so the metric is different from the intergenic field from the genomic coverage section.

rRNA

Number of aligned reads mapping to Ribosomal RNA regions. Regions are extracted from the provided annotation using the following values of the gene_type attribute:

• rRNA
• Mt_rRNA

metrics

Fractional metrics for the following read types:

mapped	mapped reads over the total number of reads
intergenic	number of reads falling in intergenic regions over the number of mapped reads
rRNA	number of reads falling in ribosomal regions over the number of mapped reads
duplicates	number of duplicate reads over the number of mapped reads