seqwho 1.0.3

Rapid and accurate classification of FASTQ(A) sequence files

SeqWho - An accruate and rapid FASTQ(A) file origin classifier

This is the official SeqWho Repository.

SeqWho is a reliable and extremely rapid program designed to determine a FASTQ(A) sequencing file identity, both source protocol and species of origin. This is accomplished using an alignment-free algorithm that leverages a Random Forest classifier that learns from biases in k-mer frequencies and repeat sequence identity. SeqWho is capable of achieving greater than 96% accuracy in its ability to classify files.

You can find the Documentation for SeqWho at: https://daehwankimlab.github.io/seqwho/

First time setup

SeqWho is written in Python 3 and we recommend using a conda environment built from the environment.yml included with SeqWho for optimal performance.

Please read https://daehwankimlab.github.io/seqwho/manual/ for more details.

Download pre-trained SeqWho index

Species	Libraries	Index
Human, Mouse	Amplicon, ChIP-Seq, WGS, WES, miRNA-Seq, RNA-Seq, Bisulfite-Seq, DNase-Seq, ATAC-Seq	Index(md5sum) Training File List Testing File List
Human, Mouse, Rattus norvegicus	ChIP-Seq, WGS, RNA-Seq	Index(md5sum) Training File List Testing File List
Human, Mouse, Arabidopsis thaliana, Caenorhabditis elegans, Drosophila melanogaster, Saccharomyces cerevisiae	ChIP-Seq, WGS, RNA-Seq	Index(md5sum) Training File List Testing File List

Current release

v1.0.3 - Added option to select number of reads drawn from files during model building

v1.0.2 - Removed extra commas in some fields to facilitate CSV conversion

v1.0.1 - Addition of test files and scripts

v1.0.0 - Initial public release