Reads Binning Improves Alignment-Free Metagenome Comparison

Abstract

Comparing metagenomic samples is a critical step in understanding the relationships among microbial communities. Recently, next-generation sequencing (NGS) technologies have produced a massive amount of short reads data for microbial communities from different environments. The assembly of these short reads can, however, be time-consuming and challenging. In addition, alignment-based methods for metagenome comparison are limited by incomplete genome and/or pathway databases. In contrast, alignment-free methods for metagenome comparison do not depend on the completeness of genome or pathway databases. Still, the existing alignment-free methods, d2S and d2*, which model k-tuple patterns using only one Markov chain for each sample, neglect the heterogeneity within metagenomic data wherein potentially thousands of types of microorganisms are sequenced. To address this imperfection in d2S and d2*, we organized NGS sequences into different reads bins and constructed several corresponding Markov models. Next, we modified the definition of our previous alignment-free methods, d2S and d2*, to make them more compatible with a scheme of analysis which uses the proposed reads bins. We then used two simulated and three real metagenomic datasets to test the effect of the k-tuple size and Markov orders of background sequences on the performance of these de novo alignment-free methods. For dependable comparison of metagenomic samples, our newly developed alignment-free methods with reads binning outperformed alignment-free methods without reads binning in detecting the relationship among microbial communities, including whether they form groups or change according to some environmental gradients.

geneRFinder: gene finding in distinct metagenomic data complexities

“…The first group refers to GC content features, these features are classic ones in gene prediction. The second group refers to k-mer features, these features are widely used in other branches of Bioinformatics such as assembly [ 25 ] and binning [ 30 ], but still little explored in gene prediction problems.

Fig.

…”

KmerGO: A Tool to Identify Group-Specific Sequences With k-mers

“…Short k -mer ( k < 15) based measures, such as , and CVtree , calculate dissimilarity between sequences or high-throughput sequencing samples ( Jiang et al, 2012 ; Liao et al, 2016 ; Song et al, 2019 ) using the global statistical models. Based on long k -mers ( k > 21), Mash ( Ondov et al, 2016 ), Skmer ( Sarmashghi et al, 2019 ), and Kmer-db ( Deorowicz et al, 2018 ) use MinHash to approximate Jaccard distance between pairwise sequences based on randomly sampled small set of k -mers.…”

geneRFinder: gene finding in distinct metagenomic data complexities

“…The first group refers to GC content features, these features are classic ones in gene prediction, being used in tools such as FragGeneScan, Prodigal and Orphelia. The second group refers to k-mer features, these features are widely used in other branches of Bioinformatics such as assembly [30] and binning [31], but still little explored in gene prediction problems. The feature importance index was calculated according to importance method of the Caret package [32] and, as Figure 3 presents the sequence length as the most important one, followed by k-mer features, having more than 80% importance index.…”