The impact of contaminants on the accuracy of genome skimming and the effectiveness of exclusion read filters

Abstract

The ability to detect the identity of a sample obtained from its environment is a cornerstone of molecular ecological research. Thanks to the falling price of shotgun sequencing, genome skimming, the acquisition of short reads spread across the genome at low coverage, is emerging as an alternative to traditional barcoding. By obtaining far more data across the whole genome, skimming has the promise to increase the precision of sample identification beyond traditional barcoding while keeping the costs manageable. While methods for assembly‐free sample identification based on genome skims are now available, little is known about how these methods react to the presence of DNA from organisms other than the target species. In this paper, we show that the accuracy of distances computed between a pair of genome skims based on k‐mer similarity can degrade dramatically if the skims include contaminant reads; i.e., any reads originating from other organisms. We establish a theoretical model of the impact of contamination. We then suggest and evaluate a solution to the contamination problem: Query reads in a genome skim against an extensive database of possible contaminants (e.g., all microbial organisms) and filter out any read that matches. We evaluate the effectiveness of this strategy when implemented using Kraken‐II, in detailed analyses. Our results show substantial improvements in accuracy as a result of filtering but also point to limitations, including a need for relatively close matches in the contaminant database.

CONSULT: Accurate contamination removal using locality-sensitive hashing

“…Kraken reference libraries were built without masking low-complexity sequences, but using default settings otherwise. We note that (27) found defaults were the most effective setting for contamination removal.…”

“…Similar to (27), we first evaluated the ability of CONSULT to find a match when the query is within a range of phylogenetic distances to the closest species present in a database. To control the proximity of the query to its closest match in the reference library, we selected 100 genomes from TOL such that their distances to their closest species in the tree uniformly covered a broad range of [0.0-0.3).…”

“…Such contamination has the potential to reduce accuracy of distances estimated from genome skims. Using theoretical modeling and experimental studies, (27) have shown that contamination can lead to over and underestimation of distances between genome skims using assembly-free methods such as Skmer.…”

“…For instance, the phylum level classification lacks sensitivity when tested on novel data (42). We recently showed (27) that even at the domain-level, the sensitivity of the leading method Kraken-II (39) degrades dramatically as the distance to the closest match in the database increases above ≈8% demonstrating that even leading metagenomic methods have serious limitations for contamination removal. The limited sensitivity of methods has spurred the development of many reference sets (46, 47, 48, 49), including recent whole-genome databases with up to 25000 genomes (50, 51, 52).…”

Estimating repeat spectra and genome length from low-coverage genome skims with RESPECT

“…in the reference databases26 . We discarded 10 samples because > 40% of reads (after removing adapters) were either duplicates of other reads, or came from external DNA sources (Supplementary Table S1).…”