Binning sequences using very sparse labels within a metagenome

Abstract

Background: In metagenomic studies, a process called binning is necessary to assign contigs that belong to multiple species to their respective phylogenetic groups. Most of the current methods of binning, such as BLAST, k-mer and PhyloPythia, involve assigning sequence fragments by comparing sequence similarity or sequence composition with already-sequenced genomes that are still far from comprehensive. We propose a semi-supervised seeding method for binning that does not depend on knowledge of completed genomes. Instead, it extracts the flanking sequences of highly conserved 16S rRNA from the metagenome and uses them as seeds (labels) to assign other reads based on their compositional similarity.

Results: The proposed seeding method is implemented on an unsupervised Growing Self-Organising Map (GSOM), and called Seeded GSOM (S-GSOM). We compared it with four well-known semi-supervised learning methods in a preliminary test, separating random-length prokaryotic sequence fragments sampled from the NCBI genome database. We identified the flanking sequences of the highly conserved 16S rRNA as suitable seeds that could be used to group the sequence fragments according to their species. S-GSOM showed superior performance compared to the semi-supervised methods tested. Additionally, S-GSOM may also be used to visually identify some species that do not have seeds. The proposed method was then applied to simulated metagenomic datasets using two different confidence threshold settings and compared with PhyloPythia, k-mer and BLAST. At the reference taxonomic level Order, S-GSOM outperformed all k-mer and BLAST results and showed comparable results with PhyloPythia for each of the corresponding confidence settings, where S-GSOM performed better than PhyloPythia in the >/= 10 reads datasets and comparable in the > or = 8 kb benchmark tests.

Conclusion: In the task of binning using semi-supervised learning methods, results indicate S-GSOM to be the best of the methods tested. Most importantly, the proposed method does not require knowledge from known genomes and uses only very few labels (one per species is sufficient in most cases), which are extracted from the metagenome itself. These advantages make it a very attractive binning method. S-GSOM outperformed the binning methods that depend on already-sequenced genomes, and compares well to the current most advanced binning method, PhyloPythia.

Figure 1

An example for preparing unlabelled input vectors and seeds. Unlabelled input vectors and seeds are prepared by avoiding the RNA sequences.

Figure 2

The S-GSOM algorithm. (a) Schematic diagram of the clustering process of S-GSOM; (b) The pseudo code for node assigning process in S-GSOM.

Figure 3

An overview of binning process using S-GSOM.

Figure 4

Identification of an appropriate Clustering Percentage (CP). Five datasets for each of 5, 10 and 20 species are randomly sampled. The averages of S-GSOM's clustering performance for the datasets are plotted against Clustering Percentage (CP) values. A trend of decreasing in clustering performance with increasing CP can be noted. A compromised value of CP = 55% is marked where both the number of assigned nodes and clustering performance are high.

Figure 5

Resulted GSOM maps of randomly sampled species. The figure illustrates the GSOM results of clustering sequence fragments according to species: (a) 10Sp_Set1, (b) 20Sp_Set1 and (c) 40Sp_Set1. Each hexagon represents a single node. If it only contains a single species, it is displayed in a colour that uniquely identifies the species. A node without a letter means that there is no sample located in it. The grey node represents two or more species in the node and the number of species is displayed on the node.

Figure 6

Illustration of exploring an unseeded cluster. (a) The 5-species S-GSOM map. The seeded nodes are shown with unique colours and labels. Nodes in charcoal colour represent nodes that will be assigned when CP = 27% and dark grey nodes at CP = 55%, light grey at CP = 77%, and white at CP = 100%. (b) Inter-node distance map with nodes assigned at CP = 55%.