karel-brinda / rase

The core RASE package

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

RASE - Resistance-Associated Sequence Elements

This repository contains the core RASE package. Other components of the software include the RASE pipeline and the RASE DB skeleton. For more information, see the RASE paper and the RASE supplementary materials.

Content

Introduction

This repository contains the core RASE software for rapid inference of antibiotic resistance and susceptibility using genomic neighbor typing.

Installation

  1. Setup an environment

  2. Install RASE using PIP

    pip install --upgrade git+https://github.com/c2-d2/rase

File formats

Prediction output (timeline)

Tab-separated text file with the following columns:

column description
datetime datetime of the read
reads number of processed reads
kbps number of processed bps (thousands)
kkmers number of matched k-mers (thousands)
ks k-mer score (proportion of matched k-mers)
ls lineage score
ls_pass lineage score interpretation, 1=passing 0=failing
ln, alt_ln predicted and alternative lineage
bm, bm_{feature} best-matching strain (nearest neighbor) and its features
{ant}_ss susceptibility score for the antibiotic {ant}
{ant}_pred prediction (score interpretation): S=susceptible, R=non-susceptible, S! and R!=low confidence calls
{ant}_bm resistance information for the best match, format: cat (mic)

See an example file.

Prediction output (snapshot)

Tab-separated text file with the following columns:

column description
taxid taxid of a database strain, _unassigned_ for reads without any k-mer matches with the database
lineage lineage
weight weight (cumulative "number of k-mers matches divided by the the number of best matches")
weight_norm normalized weight
length cumulative "read length divided by number of matches"
count cumulative "count divided by number of matches"

See an example file.

RASE DB metadata

Tab-separated text file with the following columns:

column description
taxid taxid of a database strain
lineage lineage
order order for plotting
{feature} arbitrary features (e.g., serotype or MLST)
{ant}_mic original MIC string (e.g., <0.03)
{ant}_int extracted MIC interval (0-0.03)
{ant}_cat resistance category (S/R/s/r)

See an example file. Metadata files are generated in dedicated repositories (see RASE DB skeleton, S. pneumoniae RASE DB, and N. gonorrhoeae RASE DB).

RASE/BAM

File following the SAM/BAM specification, containing the following information:

field description
QNAME read name (ideally with the timestamp encoded)
FLAG 4 for unassigned, 0 otherwise
RNAME strain taxid or name of an internal node
ln:i: read length
h1:i: number of k-mer matches at this node

Related repositories

  • RASE supplementary. Supplementary Materials for the RASE paper, including figures, tables, experiments, and links to other related repositories.
  • RASE DB skeleton. Skeleton for creating novel RASE databases.

License

MIT.

Contact

Karel Brinda <kbrinda@hsph.harvard.edu>

About

The core RASE package

License:MIT License


Languages

Language:Python 67.4%Language:R 29.3%Language:Makefile 2.1%Language:Shell 1.3%