BCT

de Bruijn graph based corrector for transcriptomic data

Installation:

./install.sh -t 8

Will install bct using 8 thread for the compilation

Usage:

Basic usage:

./Bct.py -u reads.fa -o working_directory -t core_number

More agressive filtering:

./Bct.py -u reads.fa -o working_directory -t core_number -S abundance_threshold

With this option the graph will contain less errors but regions seen less than S times may be lost

Disable poly A tail cleaning:

./Bct.py -u reads.fa -o working_directory -t core_number -S abundance_threshold -c 0

Use this option if you do not have polyA tails in your dataset, in meta-genomic for example.

The polyA tail is HIGHLY recommanded for transriptomic data.

Other options are advanced parameters, change them at your own risk

Option:

-h display help

You are lost

-u read file

Your input reads in fasta/multifasta/fastq file gziped or not

If you have multiple dataset please concatenate them

I assume fastq file will contain ".fq" or ".FQ" or ".fastq" or ".FASTQ" if not please add -q to the command line

-o output directory

The cleaned graph, the logs and the corrected reads will be put in this directory (default .)

-t thread number

The number of threads we can run in parallel (default 20)

-k kmer size

The kmer size used to build the graph (default 31)

-s kmer abundance threshold

kmer seen less than s time will be lost (default 2)

-S unitig abundance threshold

unitig seen less than S time will be lost (default 2)

-i anchors subsampling

Index 1 out of i anchors to reduce memory usage (default all)

-n repeated anchors

Anchors seen more than n time are not indexed (default 8)

-d debug mode

Will print the command line used by BCT