Machine learning on the road to unlocking microbiota's potential for boosting immune checkpoint therapy.

The intestinal microbiota is a complex and diverse ecological community that fulfills multiple functions and substantially impacts human health. Despite its plasticity, unfavorable conditions can cause perturbations leading to so-called dysbiosis, which have been connected to multiple diseases. Unfortunately, understanding the mechanisms underlying the crosstalk between those microorganisms and their host is proving to be difficult. Traditionally used bioinformatic tools have difficulties to fully exploit big data generated for this purpose by modern high throughput screens. Machine Learning (ML) may be a potential means of solving such problems, but it requires diligent application to allow for drawing valid conclusions. This is especially crucial as gaining insight into the mechanistic basis of microbial impact on human health is highly anticipated in numerous fields of study. This includes oncology, where growing amounts of studies implicate the gut ecosystems in both cancerogenesis and antineoplastic treatment outcomes. Based on these reports and first signs of clinical benefits related to microbiota modulation in human trials, hopes are rising for the development of microbiome-derived diagnostics and therapeutics. In this mini-review, we're inspecting analytical approaches used to uncover the role of gut microbiome in immune checkpoint therapy (ICT) with the use of shotgun metagenomic sequencing (SMS) data.

Host Factors Associated with Gut Mycobiome Structure.

Recent studies revealed a significant role of the gut fungal community in human health. Here, we investigated the content and variation of gut mycobiota among subjects from the European population. We explored the interplay between gut fungi and various host-related sociodemographic, lifestyle, health, and dietary factors. The study included 923 participants. Fecal DNA samples were analyzed by whole-metagenome high-throughput sequencing. Subsequently, fungi taxonomic profiles were determined and accompanied by computational and statistical analyses of the association with 53 host-related factors. Fungal communities were characterized by a high prevalence of Saccharomyces, Candida, and Sporisorium. Ten factors were found to correlate significantly with the overall mycobiota variation. Most were diet related, including the consumption of chips, meat, sodas, sweetening, processed food, and alcohol, followed by age and marital status. Differences in α- and/or ß-diversity were also reported for other factors such as body mass index (BMI), job type, autoimmunological diseases, and probiotics. Differential abundance analysis revealed fungal species that exhibited different patterns of changes under specific conditions. The human gut mycobiota is dominated by yeast, including Saccharomyces, Malassezia, and Candida. Although intervolunteer variability was high, several fungal species persisted across most samples, which may be evidence that a core gut mycobiota exists. Moreover, we showed that host-related factors such as diet, age, and marital status influence the variability of gut mycobiota. To our knowledge, this is the first large and comprehensive study of the European cohort in terms of gut mycobiota associations with such an extensive and differentiated host-related set of factors. IMPORTANCE The human gut is inhabited by many organisms, including bacteria and fungi, that may affect human health. However, research on human gut mycobiome is still rare. Moreover, the large European-based cohort study is missing. Here, we analyzed the first large European cohort in terms of gut mycobiota associations with a differentiated host-related set of factors. Our results showed that chips, meat, sodas, sweetening, processed food, beer, alcohol consumption, age, and marital status were associated with the variability of gut mycobiota. Moreover, our analysis revealed changes in abundances at the fungal species level for many investigated factors. Our results can suggest potentially valuable paths for further, narrowly focused research on gut mycobiome and its impact on human health. In the coming era of gut microbiome-based precision medicine, further research into the relationship between different mycobial structures and host-related factors may result in new preventive approaches or therapeutic procedures.

Deep embeddings to comprehend and visualize microbiome protein space.

Understanding the function of microbial proteins is essential to reveal the clinical potential of the microbiome. The application of high-throughput sequencing technologies allows for fast and increasingly cheaper acquisition of data from microbial communities. However, many of the inferred protein sequences are novel and not catalogued, hence the possibility of predicting their function through conventional homology-based approaches is limited, which indicates the need for further research on alignment-free methods. Here, we leverage a deep-learning-based representation of proteins to assess its utility in alignment-free analysis of microbial proteins. We trained a language model on the Unified Human Gastrointestinal Protein catalogue and validated the resulting protein representation on the bacterial part of the SwissProt database. Finally, we present a use case on proteins involved in SCFA metabolism. Results indicate that the deep learning model manages to accurately represent features related to protein structure and function, allowing for alignment-free protein analyses. Technologies that contextualize metagenomic data are a promising direction to deeply understand the microbiome.

The standardisation of the approach to metagenomic human gut analysis: from sample collection to microbiome profiling.

In recent years, the number of metagenomic studies increased significantly. Wide range of factors, including the tremendous community complexity and variability, is contributing to the challenge in reliable microbiome community profiling. Many approaches have been proposed to overcome these problems making hardly possible to compare results of different studies. The significant differences between procedures used in metagenomic research are reflected in a variation of the obtained results. This calls for the need for standardisation of the procedure, to reduce the confounding factors originating from DNA isolation, sequencing and bioinformatics analyses in order to ensure that the differences in microbiome composition are of a true biological origin. Although the best practices for metagenomics studies have been the topic of several publications and the main aim of the International Human Microbiome Standard (IHMS) project, standardisation of the procedure for generating and analysing metagenomic data is still far from being achieved. To highlight the difficulties in the standardisation of metagenomics methods, we thoroughly examined each step of the analysis of the human gut microbiome. We tested the DNA isolation procedure, preparation of NGS libraries for next-generation sequencing, and bioinformatics analysis, aimed at identifying microbial taxa. We showed that the homogenisation time is the leading factor impacting sample diversity, with the recommendation for a shorter homogenisation time (10 min). Ten minutes of homogenisation allows for better reflection of the bacteria gram-positive/gram-negative ratio, and the obtained results are the least heterogenous in terms of beta-diversity of samples microbial composition. Besides increasing the homogenisation time, we observed further potential impact of the library preparation kit on the gut microbiome profiling. Moreover, our analysis revealed that the choice of the library preparation kit influences the reproducibility of the results, which is an important factor that has to be taken into account in every experiment. In this study, a tagmentation-based kit allowed for obtaining the most reproducible results. We also considered the choice of the computational tool for determining the composition of intestinal microbiota, with Kraken2/Bracken pipeline outperforming MetaPhlAn2 in our in silico experiments. The design of an experiment and a detailed establishment of an experimental protocol may have a serious impact on determining the taxonomic profile of the intestinal microbiome community. Results of our experiment can be helpful for a wide range of studies that aim to better understand the role of the gut microbiome, as well as for clinical purposes.