What is a healthy microbiome? The pursuit of this and many related questions, especially in light of the recently recognized microbial component in a wide range of diseases has sparked a surge in metagenomic studies. They are often not simply attributable to a single pathogen but rather are the result of complex ecological processes. Relatedly, the increasing DNA sequencing depth and number of samples in metagenomic case-control studies enabled the applicability of powerful statistical methods, e.g. Machine Learning approaches. For the latter, the feature space is typically shaped by the relative abundances of operational taxonomic units, as determined by cost-effective phylogenetic marker gene profiles. While a substantial body of microbiome/microbiota research involves unsupervised and supervised Machine Learning, very little attention has been put on feature selection and engineering. We here propose the first algorithm to exploit phylogenetic hierarchy (i.e. an all-encompassing taxonomy) in feature engineering for microbiota classification. The rationale is to exploit the often mono- or oligophyletic distribution of relevant (but hidden) traits by virtue of taxonomic abstraction. The algorithm is embedded in a comprehensive microbiota classification pipeline, which we applied to a diverse range of datasets, distinguishing healthy from diseased microbiota samples. We demonstrate substantial improvements over the state-of-the-art microbiota classification tools in terms of classification accuracy, regardless of the actual Machine Learning technique while using drastically reduced feature spaces. Moreover, generalized features bear great explanatory value: they provide a concise description of conditions and thus help to provide pathophysiological insights. Indeed, the automatically and reproducibly derived features are consistent with previously published domain expert analyses.

Scientific research is shedding light on the interaction of the gut microbiome with the human host and on its role in human health. Existing machine learning methods have shown great potential in discriminating healthy from diseased microbiome states. Most of them leverage shotgun metagenomic sequencing to extract gut microbial species-relative abundances or strain-level markers. Each of these gut microbial profiling modalities showed diagnostic potential when tested separately; however, no existing approach combines them in a single predictive framework. Here, we propose the Multimodal Variational Information Bottleneck (MVIB), a novel deep learning model capable of learning a joint representation of multiple heterogeneous data modalities. MVIB achieves competitive classification performance while being faster than existing methods. Additionally, MVIB offers interpretable results. Our model adopts an information theoretic interpretation of deep neural networks and computes a joint stochastic encoding of different input data modalities. We use MVIB to predict whether human hosts are affected by a certain disease by jointly analysing gut microbial species-relative abundances and strain-level markers. MVIB is evaluated on human gut metagenomic samples from 11 publicly available disease cohorts covering 6 different diseases. We achieve high performance (0.80 < ROC AUC < 0.95) on 5 cohorts and at least medium performance on the remaining ones. We adopt a saliency technique to interpret the output of MVIB and identify the most relevant microbial species and strain-level markers to the model’s predictions. We also perform cross-study generalisation experiments, where we train and test MVIB on different cohorts of the same disease, and overall we achieve comparable results to the baseline approach, i.e. the Random Forest. Further, we evaluate our model by adding metabolomic data derived from mass spectrometry as a third input modality. Our method is scalable with respect to input data modalities and has an average training time of < 1.4 seconds. The source code and the datasets used in this work are publicly available.

Surveillance methods of circulating antibiotic resistance genes (ARGs) are of utmost importance in order to tackle what has been described as one of the greatest threats to humanity in the 21^st century. In order to be effective, these methods have to be accurate, quickly deployable, and scalable. In this study, we compare metagenomic shotgun sequencing (TruSeq DNA sequencing) of wastewater samples with a state-of-the-art PCR-based method (Resistomap HT-qPCR) on four wastewater samples that were taken from hospital, industrial, urban and rural areas. ARGs that confer resistance to 11 antibiotic classes have been identified in these wastewater samples using both methods, with the most abundant observed classes of ARGs conferring resistance to aminoglycoside, multidrug-resistance (MDR), macrolide-lincosamide-streptogramin B (MLSB), tetracycline and beta-lactams. In comparing the methods, we observed a strong correlation of relative abundance of ARGs obtained by the two tested methods for the majority of antibiotic classes. Finally, we investigated the source of discrepancies in the results obtained by the two methods. This analysis revealed that false negatives were more likely to occur in qPCR due to mutated primer target sites, whereas ARGs with incomplete or low coverage were not detected by the sequencing method due to the parameters set in the bioinformatics pipeline. Indeed, despite the good correlation between the methods, each has its advantages and disadvantages which are also discussed here. By using both methods together, a more robust ARG surveillance program can be established. Overall, the work described here can aid wastewater treatment plants that plan on implementing an ARG surveillance program.

Comprehensive mapping of environmental microbiomes in terms of their compositional features remains a great challenge in understanding the microbial biosphere of the Earth. It bears promise to identify the driving forces behind the observed community patterns and whether community assembly happens deterministically. Advances in Next Generation Sequencing allow large community profiling studies, exceeding sequencing data output of conventional methods in scale by orders of magnitude. However, appropriate collection systems are still in a nascent state. We here present a database of 20,427 diverse environmental 16S rRNA profiles from 2,426 independent studies, which forms the foundation of our meta-analysis. We conducted a sample size adaptive all-against-all beta diversity comparison while also respecting phylogenetic relationships of Operational Taxonomic Units(OTUs). After conventional hierarchical clustering we systematically test for enrichment of Environmental Ontology terms and their abstractions in all possible clusters. This post-hoc algorithm provides a novel formalism that quantifies to what extend compositional and semantic similarity of microbial community samples coincide. We automatically visualize significantly enriched subclusters on a comprehensive dendrogram of microbial communities. As a result we obtain the hitherto most differentiated and comprehensive view on global patterns of microbial community diversity. We observe strong clusterability of microbial communities in ecosystems such as human/mammal-associated, geothermal, fresh water, plant-associated, soils and rhizosphere microbiomes, whereas hypersaline and anthropogenic samples are less homogeneous. Moreover, saline samples appear less cohesive in terms of compositional properties than previously reported. We here set out to map the entirety of available environmental microbiomes in order to discover the underlying compositional characteristics. For us it is intriguing to see which environmental factors influence the assembly of microbiomes. We collected many diverse environmental samples and annotated them with a restricted, yet structured set of ecosystem terms. We then cluster all samples and automatically detect which ecosystems cluster together. The resulting map provides an overview of relatedness between microbial communities from all types of ecosystems, their global patterns of diversity, i.e. their compositional similarities and differences. The comprehensive structure of relations provides insights to habitat adaptation and assembly rules. Finally, utilizing the background database of samples we can now put new samples into an environmental context.

Background: For some understudied populations, genotype data is minimal for genotype-phenotype prediction. However, we can use the data of some other large populations to learn about the disease-causing SNPs and use that knowledge for the genotype-phenotype prediction of small populations. This manuscript illustrated that transfer learning is applicable for genotype data and genotype-phenotype prediction. Results: Using HAPGEN2 and PhenotypeSimulator, we generated eight phenotypes for 500 cases/500 controls (CEU, large population) and 100 cases/100 controls (YRI, small populations). We considered 5 (4 phenotypes) and 10 (4 phenotypes) different risk SNPs for each phenotype to evaluate the proposed method. The improved accuracy with transfer learning for eight different phenotypes was between 2 and 14.2 percent. The two-tailed p-value between the classification accuracies for all phenotypes without transfer learning and with transfer learning was 0.0306 for five risk SNPs phenotypes and 0.0478 for ten risk SNPs phenotypes. Conclusion: The proposed pipeline is used to transfer knowledge for the case/control classification of the small population. In addition, we argue that this method can also be used in the realm of endangered species and personalized medicine. If the large population data is extensive compared to small population data, expect transfer learning results to improve significantly. We show that Transfer learning is capable to create powerful models for genotype-phenotype predictions in large, well-studied populations and fine-tune these models to populations were data is sparse.

For optimal pancreatic cancer treatment, early and accurate diagnosis is vital. Blood-derived biomarkers and genetic predispositions can contribute to early diagnosis, but they often have limited accuracy or applicability. Here, we seek to exploit the synergy between them by combining the biomarker CA19-9 with RNA-based variants. We use deep sequencing and deep learning to improve differentiating pancreatic cancer and chronic pancreatitis. We obtained samples of nucleated cells found in peripheral blood from 268 patients suffering from resectable, non-resectable pancreatic cancer, and chronic pancreatitis. We sequenced RNA with high coverage and obtained millions of variants. The high-quality variants served as input together with CA19-9 values to deep learning models. Our model achieved an area under the curve (AUC) of 96% in differentiating resectable cancer from pancreatitis using a test cohort. Moreover, we identified variants to estimate survival in resectable cancer. We show that the blood transcriptome harbours variants, which can substantially improve noninvasive clinical diagnosis.