.center[ # Latent Structure Laboratory ] .pull-left[ .pull-right[ <img src="figures/microbiome.png"/ width=400> ] ] .pull-right[ Bacteria may not build cities or have interesting social lives, but they will be here when the Sun explodes. -- Bill Bryson in _A Short History of Nearly Everything_ ] .center[ ### UW Madison Statistics Lightning Talks .large[Kris Sankaran | [krisrs1128.github.io/LSLab](krisrs1128.github.io/LSLab) | 16 September 2021] ] --- ### Problem Solving in Multi-Omics Our lab focuses on statistical methods for analyzing multi-omics, and especially microbiome, data: .pull-left[ * We are motivated by interdisciplinary projects with microbiologists, ecologists, and psychologists * We synthesize and benchmark statistical approaches to solve data analysis problems within these collaborations ] .pull-right[ <img src="figures/spatial_microbiome.png" width = "400"/> ] --- ### Project 1: Multiscale Topic Alignment * We were motivated by a project about bacterial vaginosis * The literature collapses microbiomes into 5 “community state types,” but we noticed subtle transitions in participants with the disease * To addressed this, we developed meta-algorithms for visualizing ensembles of topic models trained across several scales .center[ <img src="https://github.com/krisrs1128/LSLab/raw/main/assets/img/alto.png" width = 750/> ] --- ### Project 1: Multiscale Topic Alignment * We found that properties of the alignment graph are useful for count data analysis more generally - Can help quantify the degree of mixing across clusters - Can help highlight mutually exclusive species * We wrapped our algorithms in the R package [`alto`](lasy.github.io/alto) .center[ <img src="https://github.com/krisrs1128/talks/blob/master/2022/20220407/figure/refinement_lda.png?raw=true" width="550"/> ] --- ### Project 2: Meditation and the Microbiome * The Center for Healthy Minds at UW - Madison has had a long history of studying the effects of meditation on well-being * Motivated by increasing evidence of a gut-brain axis, they now also collect microbiome samples .center[ <img src="figures/brain-scan-image-monk-WEB.jpeg" width="500"/> ] --- ### Project 2: Meditation and the Microbiome .pull-left[ * We are supporting a longitudinal RCT that is gathering data from 3000 participants (for reference, my PhD thesis analyzed three subjects...) * We have adapted concepts from causal mediation analysis to relate sequencing with psychological survey data ] .pull-right[ <img src="figures/counterfactual-mediation1.svg" width="550"/> An approach to microbiome mediation analysis designed by Xinran Miao and Hanying Jiang. ] --- ### Techniques A few ideas recur throughout our research, * Generative models: These allow us to encode latent structure and integrate across data sources and scales * Data visualization: Good representations enable hypothesis generation, model criticism, and communication * Reusable software: We release our methods as accessible R packages .center[ <img src="figures/combined_example.png" width="650"/> Visualizations and code by Zhuoyan Xu and Tinghui Xu in [2 - 4]. ] --- ### Reaching Out * You can learn more at [krisrs1128.github.io/LSLab](krisrs1128.github.io/LSLab). * I encourage you to reach out for any reason -- I'm always happy to talk about statistics. * Email: ksankaran@wisc.edu --- ### References [1] Julia Fukuyama, Kris Sankaran, Laura Symul, Multiscale analysis of count data through topic alignment, Biostatistics, 2022;, kxac018, (https://doi.org/10.1093/biostatistics/kxac018) [2] Kris Sankaran and Susan P. Holmes. Generative Models: An Interdisciplinary Perspective. To appear in Annual Review of Statistics and its Applications, 2022. (https://doi.org/10.48550/arXiv.2208.06011) [3] Tinghui Xu and Kris Sankaran. Interactive visualization of spatial omics neighborhoods [version 1; peer review: awaiting peer review]. F1000Research, 2022, 11:799 (https://doi.org/10.12688/f1000research.122113.1) [4] Zhuoyan Xu and Kris Sankaran. Spatial transcriptomics dimensionality reduction using wavelet bases [version 1; peer review: awaiting peer review]. F1000Research 2022, 11:1033 (https://doi.org/10.12688/f1000research.122775.1)