Search for: All records

Recurrent neural networks with memory and attention mechanisms are widely used in natural language processing because they can capture short and long term sequential information for diverse tasks. We propose an integrated deep learning model for microbial DNA sequence data, which exploits convolutional neural networks, recurrent neural networks, and attention mechanisms to predict taxonomic classifications and sample-associated attributes, such as the relationship between the microbiome and host phenotype, on the read/sequence level. In this paper, we develop this novel deep learning approach and evaluate its application to amplicon sequences. We apply our approach to short DNA reads and full sequences of 16S ribosomal RNA (rRNA) marker genes, which identify the heterogeneity of a microbial community sample. We demonstrate that our implementation of a novel attention-based deep network architecture, Read2Pheno , achieves read-level phenotypic prediction. Training Read2Pheno models will encode sequences (reads) into dense, meaningful representations: learned embedded vectors output from the intermediate layer of the network model, which can provide biological insight when visualized. The attention layer of Read2Pheno models can also automatically identify nucleotide regions in reads/sequences which are particularly informative for classification. As such, this novel approach can avoid pre/post-processing and manual interpretation required with conventional approaches to microbiome sequence classification. We further show, as proof-of-concept, that aggregating read-level information can robustly predict microbial community properties, host phenotype, and taxonomic classification, with performance at least comparable to conventional approaches. An implementation of the attention-based deep learning network is available at https://github.com/EESI/sequence_attention (a python package) and https://github.com/EESI/seq2att (a command line tool). 

Abstract Facultative, heritable endosymbionts are found at intermediate prevalence within most insect species, playing frequent roles in their hosts’ defence against environmental pressures. Focusing onHamiltonella defensa, a common bacterial endosymbiont of aphids, we tested the hypothesis that such pressures impose seasonal balancing selection, shaping a widespread infection polymorphism. In our studied pea aphid (Acyrthosiphon pisum) population,Hamiltonellafrequencies ranged from 23.2% to 68.1% across a six‐month longitudinal survey. Rapid spikes and declines were often consistent across fields, and we estimated that selection coefficients forHamiltonella‐infected aphids changed sign within this field season. Prior laboratory research suggested antiparasitoid defence as the majorHamiltonellabenefit, and costs under parasitoid absence. While a prior field study suggested these forces can sometimes act as counter‐weights in a regime of seasonal balancing selection, our present survey showed no significant relationship between parasitoid wasps andHamiltonellaprevalence. Field cage experiments provided some explanation: parasitoids drove modest ~10% boosts toHamiltonellafrequencies that would be hard to detect under less controlled conditions. They also showed thatHamiltonellawas not always costly under parasitoid exclusion, contradicting another prediction. Instead, our longitudinal survey – and two overwintering studies – showed temperature to be the strongest predictor ofHamiltonellaprevalence. Matching some prior lab discoveries, this suggested that thermally sensitive costs and benefits, unrelated to parasitism, can shapeHamiltonelladynamics. These results add to a growing body of evidence for rapid, seasonal adaptation in multivoltine organisms, suggesting that such adaptation can be mediated through the diverse impacts of heritable bacterial endosymbionts.