Computational microbiome research in population cohort studies
Data-intensive microbiome research critically relies on the open development, benchmarking, and application of computational methods from sequence-level analyses to downstream ecological and epidemiological modeling. This project will focus on the design of new computational strategies and open data science techniques for the integration, analysis and modeling of deeply sequenced fecal microbiomes in the context of population health research. Population studies of the human microbiome can elucidate the mediating role of microbiomes between e.g. cardiovascular disease risk or brain health, demographics and lifestyle.
Our track record includes the development of widely used statistical and machine learning techniques and open data science methods for microbiome studies. We carry out methodologically oriented research on metagenome-profiled population cohorts that cover the entire life span and three generations of families, complemented by long follow-up times, detailed population register data, host phenotyping, and complementary omics from over ten thousand volunteers. The work is supported by close collaboration and networking opportunities with local and international experts in computational science, statistical ecology, and microbiome research. The balance between methods development and application can be adjusted based on the applicant’s interests and expertise.
In our group, we appreciate scientists with a PhD in a relevant computationally oriented field, or from application field with a strong methodological focus, a solid publication track record and experience in designing computational methods and workflows. Also, we cherish a multidisciplinary mindset, familiarity with command line, and fluency in at least one open data science language (e.g. R/Bioconductor, Python, Julia), which are essential tools in our group. Experience in probabilistic methods, high-performance computing environments, and open research software development are beneficial. Also, we work both independently and as a team with good communication and interpersonal skills. Good English is our spoken as well as scientific writing language.
Key words:
Open data science, Computational science, Deep-sequence data analysis, Population cohorts,
Statistical ecology, Microbiome, Metagenome
Selected publications:
- Salosensaari A et al. Taxonomic signatures of cause-specific mortality risk in human gut microbiome. Nature Communications. 2021 May 11;12(1):2671.
- Ruuskanen MO et al. Gut Microbiome Composition Is Predictive of Incident Type 2 Diabetes in a Population Cohort of 5,572 Finnish Adults. Diabetes Care. 2022 Apr 1;45(4):811-818. 3. Arani B, Nes E, Lahti L, Carpenter S & Scheffer M. Exit time as a measure of ecological resilience. Science 372(6547), 2021
- Laitinen V, Dakos V & Lahti L. Probabilistic early warning signals. Ecology & Evolution 11(20), 2021. 5. Gao Y, Şimşek Y, Gheysen E, Borman T, Li Y, Lahti L, Faust K & Garza D. miaSim: an R/Bioconductor package to easily simulate microbial community dynamics. Methods in Ecology and Evolution 14, 2023
- Khalighi M, Sommeria-Klein G, Gonze D, Faust K & Lahti L. Quantifying the impact of ecological memory on the dynamics of interacting communities. PLoS Computational Biology 18(6), 2022
Gut Microbiome and Cardiometabolic Health
The strong links between the human gut microbiome and cardiometabolic disease (CMD) have been widely studied. However, nearly all of these studies 1) have been performed using retrospective case-control samples; 2) have not assessed the mechanisms underlying these links; 3) have not replicated their results in external cohorts; and 4) have not assessed whether the gut microbiome provides any incremental predictive value over other types of omics, such as genomics or metabolomics.
Our future research project/studies will substantially expand our prior research on the epidemiology and pathophysiology of CMD. It will provide new understanding of the human microbiome, metabolome, and genome, and their relation to CMD development in humans. We will employ high-throughput metabolomic and metagenomic sequencing approaches to comprehensively map the human milieu interior across thousands of individuals from several multi-ethnic population cohorts and over a follow-up of several decades. We will then examine the association of the metagenome, hundreds of microbial-derived metabolites, and the host genome with the prospective development of CMD to improve risk prediction and to unveil new, potentially modifiable contributors to disease.
The overarching goal of this study will be to provide data on the long-term within-individual and population-level changes in the human gut microbiome over a time span of >20 years. We will also address the information gaps on the correlates of these changes and on the relation of long-term changes in the human gut microbiome with CMD.
The study design includes longitudinal and prospective components. The study samples for this proposal consist of participants three population cohorts: 1) The FINRISK 2002 cohort (N=7,237 participants with already sequenced fecal samples) was examined in 2002. The still-living 5,700 FINRISK 2002 will be invited for a stool re-sampling or an in-depth CMD health examination during 2023-2024; 2) The Healthy Finland 2023 population cohort (N=4,000 participants with fecal samples already collected). All cohorts are continuously followed up for CMD events. External replication of results will be performed in European cohorts with mixed ethnic backgrounds.
Key words:
Cardiometabolic disease, CMD pathophysiology, CMD risk prediction, Gut microbiome, Microbiome, Metabolome, Metagenome, High-throughput sequencing
In our research group, we appreciate scientists with:
- PhD or MD with experience in Biostatistics, Bioinformatics, Computational Biology, Epidemiology, Computer Science, Biology or a related field
- Relevant background knowledge including proven bioinformatics skills and a solid publication track record
- Experience in statistical data analysis, integration, and visualization
- Fluency in at least one data science language (such as Python, R)
- Ability to work both independently and in a team
- Willingness to learn new and multidisciplinary skills as needed
- Good communication and interpersonal skills and English language skills
- Good scientific writing skills
Selected publications:
- McDonald D et al. Greengenes2 unifies microbial data in a single reference tree. Nat Biotechnol. 2023 Jul 27.
- Palmu J et al. Gut microbiome and atrial fibrillation-results from a large population-based study. EBioMedicine. 2023 May;91:104583.
- Liu Y et al. Cell Metab. 2022 May 3;34(5):719-730.e4.
- Qin Y et al. Combined effects of host genetics and diet on human gut microbiota and incident disease in a single population cohort. Nat Genet. 2022 Feb;54(2):134-142.
- Ruuskanen MO et al. Gut Microbiome Composition Is Predictive of Incident Type 2 Diabetes in a Population Cohort of 5,572 Finnish Adults. Diabetes Care. 2022 Apr 1;45(4):811-818.
- Salosensaari A et al. Taxonomic signatures of cause-specific mortality risk in human gut microbiome. Nat Commun. 2021 May 11;12(1):2671.
- Koponen KK et al. Associations of healthy food choices with gut microbiota profiles. Am J Clin Nutr. 2021 Aug 2;114(2):605-616.
- Palmu J et al. Association Between the Gut Microbiota and Blood Pressure in a Population Cohort of 6953 Individuals. J Am Heart Assoc. 2020 Aug 4;9(15):e016641.