Cleanifier: A fast and lightweight k -mer based tool to remove contamination in microbial sequence data
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Summary
When working with sequence data of human microbiomes, the first step in most analysis pipelines is the removal of human contamination. There are two reasons why the removal of human reads is a critical step in microbiome data analysis. First, many countries have strict privacy and data protection guidelines for human sequencing data. This means that microbiome data containing partly human data cannot be processed outside a hospital and cannot be published. The second and from a biological view more important reason is that human contamination may cause problems in later analysis steps, such as genome assembly and binning. Current tools are either alignment-or k -mer based. In alignment based methods, all reads are aligned against the human reference and classified as human contamination based on the alignment score of the reads. However, the alignment step needs a significant amount of time. In contrast, k -mer based tools are usually faster, because they classify the reads by querying all k -mers of the reads in an efficient reference k -mer database.
We introduce Cleanifier, a fast and lightweight alignment-free tool to detect and remove human contamination. For this, we create an pangenome index of all human gapped k -mers, also called spaced seeds, that is made available online via Zenodo. If users are interesting in microbiome data from other hosts, they can alternatively build their own index using other reference genomes. In the filtering step, we query all k -mers of a read in the index and classify it as human based on the fraction of read base pairs covered by a k -mer. The human reads are either removed or written into a separate file for later analysis.
Availability and Implementation
The implementation is available via gitlab ( https://gitlab.com/rahmannlab/cleanifier ), PyPi ( https://pypi.org/project/cleanifier/ ) and Bioconda ( https://anaconda.org/bioconda/cleanifier ).
Contact
rahmann@cs.uni-saarland.de
