Summary of Workshop "Single Cell Genomics meets Data Science" - Session 2 "Spatial"

This is an AI generated summary. There may be inaccuracies.
Summarize another video · Purchase summarize.tech Premium

00:00:00 - 01:00:00

This video discusses how single cell genomics and data science can be used to study metastasis. The speaker discusses how spatial data can be used to identify co-located cell types in tissue pathology. They also show examples of how this technology can be used to study breast cancer metastasis.

00:00:00 The speaker will discuss methods for mapping the locations of individual cells in complex tissues using spatial transcriptomics and single cell data. This can be helpful for understanding tissue diversity, migration, communication, and disease.
00:05:00 In this workshop, participants were introduced to spatial transcriptomics and its potential applications. Auto location and spatial transcriptomic data deconvolution methods were discussed, and Next Generation pathology was introduced. Diseases were discussed in terms of their histological and molecular correlates, and the challenges faced in studying them using traditional methods were outlined.
00:10:00 In this workshop, different labs worked on compiling a single cell transcriptomics atlas of the human lungs and generating pathology-guided spatial transcriptomic data sets. The collective effort resulted in an automated annotation of the data, which showed that different cell types were present at different stages of lung injury.
00:15:00 In this workshop, the speaker discusses the benefits of using spatial transcriptomics to study single cells and data. They discuss how this technology can be used to study lung pathologies and discuss a few examples from the speaker's own data.
00:20:00 This workshop discusses how single cell genomics can be used to identify biomarkers for different stages of lung damage, as well as changes in cell composition over time. The outcomes of this analysis can help Pathologists better understand the progression of lung disease.
00:25:00 The presenter discusses how spatial data can be used to identify co-located cell types in tissue pathology. He also mentions how general consensus on the scale of tissue microenvironments is lacking. Pathologists are likely to consider a microenvironment to be smaller than 400 microns.
00:30:00 This video provides a introduction to single cell and spatial transcriptomics, which is being used to study metastatic environments. The speaker states that the project is finished or finalized this year, and that it is a medical and scientific reason to study metastasis. The speaker also states that metastatic sites are diverse and systematic in their locations, and that single cell and spatial transcriptomics is useful in understanding the overall disease situation.
00:35:00 The video discusses the differences between single-cell and RNA-seq data, and how spatial data is created from these two types of data. It also shows an overview of 15 metastasis samples used for spatial data analysis.
00:40:00 This workshop discusses different single cell genomics methods, compares their representations, and discusses how to perform downstream analysis.
00:45:00 The workshop discusses how single cell genomics can be used to analyze data, specifically compositional data. The workshop then demonstrates how spatial segmentation can be used to analyze metastasis samples.
00:50:00 The presenter showed examples of how single cell genomics and data science can be used to study breast cancer metastasis. They discussed how observation splitting and regionalization can be used to identify different cell types, and how differential expression analysis can be used to assess interpatient heterogeneity. Finally, they showed an example of using Taco to analyze compositional data.
00:55:00 The speaker discusses how single cell genomics and data science can be combined to improve validation of findings. They discuss how binning can be used to improve accuracy of segmentation and provide a biological example of how this would be useful.

01:00:00 - 01:30:00

In this workshop, participants learned about various tools and methods for studying single cell RNA and its regulatory networks. The presenter also introduced the Lyanna framework for integrating and comparing different methods. Finally, Daniel Dimitrovastun described some of the methods he and his team have developed for spatial transcriptomics.

01:00:00 The speaker discusses how prior knowledge can be used to improve machine learning and statistics when analyzing single cell or spatial omics data. They describe a tool called "columns" and discuss how it can be used to reduce large omics data into a smaller number of features, increase statistical power, and be more interpretable mechanistically. They also mention a tool called "projecting" which is used to assess the activity of a process of interest. They also discuss a tool called "tour de coupler" which is used to run enrichment methods on large data sets.
01:05:00 In this workshop, participants learned about various tools and methods for studying single cell RNA and its regulatory networks. The presenter also introduced the Lyanna framework for integrating and comparing different methods. Finally, Daniel Dimitrovastun described some of the methods he and his team have developed for spatial transcriptomics.
01:10:00 In this workshop, participants discussed how to use multiview machine learning to dissect interactions between different spots in spatial transcriptomic data. They applied the framework Misti to um, heart tissue, and found that the presence of a particular type of macrophages (spp1 positive macrophages) was associated with the presence of myofibroblasts. Christopher followed up this analysis by looking at the collocalization of these cells.
01:15:00 The speaker discusses the process of development, focusing on the timing of transcription and cellular diversity. They show a transcriptome map of a mouse embryo, which illustrates the reproducibility of the process. They discuss the ability to connect different time points and how this has made them a natural resource for studying optimality in transport.
01:20:00 In this workshop, participants learned how to use spatial data to infer transitions between different time points and cell types. The workshop also introduced the three tet enzymes, which are known to demethylate DNA. When these enzymes are knocked out, the methylation levels in different parts of the development embryo become markedly different, leading to severe developmental abnormalities.
01:25:00 In Session 2 of the Workshop "Single Cell Genomics meets Data Science", the knockouts developed an abnormal shape and were smaller than usual embryos. However, when the knockouts were repeated in the camera setting, they looked completely normal. This discrepancy was explained by the differences in methylation between the two settings.
01:30:00 This workshop discussed how single cell genomics and data science can work together to study the effects of genetic modifications on different tissues in an organism. The workshop also discussed how knockouts can be used to study these effects more closely.