3CAC: improving the classification of phages and plasmids in metagenomic assemblies using assembly graphs

Abstract

Bacteriophages and plasmids usually coexist with their host bacteria in microbial communities and play important roles in microbial evolution. Accurately identifying sequence contigs as phages, plasmids, and bacterial chromosomes in mixed metagenomic assemblies is critical for further unravelling their functions. Many classification tools have been developed for identifying either phages or plasmids in metagenomic assemblies. However, only two classifiers, PPR-Meta and viralVerify, were proposed to simultaneously identify phages and plasmids in mixed metagenomic assemblies. Due to the very high fraction of chromosome contigs in the assemblies, both tools achieve high precision in the classification of chromosomes but perform poorly in classifying phages and plasmids. Short contigs in these assemblies are often wrongly classified or classified as uncertain.Here we present 3CAC, a new three-class classifier that improves the precision of phage and plasmid classifications. 3CAC starts with an initial three-class classification generated by existing classifiers and further improves the classification of short contigs and contigs with low confidence classification by using proximity in the assembly graph. Evaluation on simulated metagenomes and on real human gut microbiome samples showed that 3CAC outperformed PPR-Meta and viralVerify in both precision and recall, and increased F1-score by at least 10 percentage points.

Exploring microbial functional biodiversity at the protein family level—From metagenomic sequence reads to annotated protein clusters

“…The latter has been used as the basis for DeepVirFinder ( Ren et al, 2020 ), a deep learning method that uses convolutional neural networks, capable of detecting viral signals in very short contigs (<5,000 bps). Other recently developed deep learning tools include 3CAC ( Pu and Shamir, 2022 ), a combined predictor of phages and bacterial plasmids, the bacteriophage-specific INHERIT ( Bai et al, 2022 ), virSearcher ( Liu et al, 2022 ), PHAMB ( Johansen et al, 2022 ), Seeker ( Auslander et al, 2020 ) and PhaMer ( Shang et al, 2022 ) predictors and DeepMicrobeFinder ( Hou et al, 2021 ), which classifies metagenomic contigs into five sequence classes (prokaryotic genomes, eukaryotic genomes, plasmids, prokaryotic-infecting viruses and eukaryotic-infecting viruses) with a reported accuracy of over 90% for viral contigs.…”

4CAC: 4-class classifier of metagenome contigs using machine learning and assembly graphs

“…Most of the existing classifiers take contigs as input and classify each of them independently based on its sequence. Our recent work on three-class classification demonstrated that neighboring contigs in an assembly graph are more likely to come from the same class and thus the adjacency information in the graph can be used to improve the classification [30]. Therefore, here too we exploit the assembly graph to improve the initial classification by the following two steps.…”

“…Plasmid classifiers, such as cBar [43], PlasFlow [15], PlasClass [27], and Deeplasmid [1], were developed to separate plasmids from prokaryotes. As both phages and plasmids are commonly found in microbial communities, three-class classifiers, such as PPR-Meta [7], viralVerify [2], and 3CAC [30], were proposed to simultaneously identify phages, plasmids, and prokaryotes from metagenome assemblies. On the other hand, microbial eukaryotes, such as fungi and protozoa, are integral components of natural microbial communities but were commonly ignored or misclassified as prokaryotes in most metagenome analyses.…”

“…However, even though prokaryotes, microeukaryotes, phages, and plasmids are present in most microbial communities, there are still no four-class classifiers that can simultaneously identify and distinguish all of them. Moreover, most classifiers ignore the fact that microbial communities are dominated by bacteria, and have low precision on the minor classes, such as phages, plasmids, and microeukaryotes [32, 30].…”

plASgraph - using graph neural networks to detect plasmid contigs from an assembly graph

“…The recent method 3CAC [24] introduced the idea that the classification of a contig can be improved from the knowledge of the classification of the neighbouring contigs in the assembly graph. Most current assembly programs [7, 28] output an assembly graph containing final contigs as nodes and possible connections between them supported by sequencing data as edges.…”