User-friendly bioinformatics (Monthly Informational workshop)
1 like124 views
The document discusses the integration of clinical and omics data for enhancing precision medicine and patient targeting in biomedical research. It highlights the use of machine learning in analyzing large datasets and preparing projects that connect public domain data with clinical outcomes. Additionally, the document provides information on education and training workshops aimed at improving data analysis skills in the biomedical field.
User-friendly bioinformatics (Monthly Informational workshop)
- 1. 1
- 2. 2 Elia Brodsky Co-founder and CEO Pine Biotech (pine-biotech.com) Survey: bit.ly/2KG7iEj Georgetown
- 3. 3 Overview Education & Training Research Support Survey: bit.ly/2KG7iEj
- 4. 4
- 5. 5 Patients Most important clinical features Normalizedclinicalfeatures Missing Values Group-to-group associations (bipartite network) OmicsFeatures PREDICTION Enrichment of Clinical Data Clinical characteristics of patients that are of most interest can be predicted using other clinical features. This type of group-to-group association network can be prepared within a normalized set of clinical features across patients, thus identifying patients at risk or selecting patients for specific data to be generated. Validated associations can become a part of a clinical application used regularly to select patients or match patients to requirements. Integration of Omics Data Omics data represents a feature-rich resource that can be used for discovery of underlying biology linked to a particular condition, disease stage, co-morbidities and population stratification. Integrating the omics data into a usable, normalized resource and linking it to clinical outcomes can enable precision targeting of patients and more precise treatment for complex diseases. Machine Learning Applications
- 6. 6 Exploring Ideas Experiment Design Research and Development Data is playing a key role in all stages of R&D:
- 7. the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus repository (GEO) alone contains 80,985 public datasets, spanning hundreds of tissue types in thousands of organisms Omics Data: What’s out there? 7
- 8. 8 Education & Training Research Support Dataset: bit.ly/2vNovGC
- 9. 9
- 10. 10 Online resources Curated Datasets User-friendly Tools
- 11. 11 The T-BioInfo Platform is at the core of this experience Server.t-bio.info
- 12. 12
- 13. 13
- 14. 14
- 15. INDEPENDENT PROJECTS (ONCOLOGY SPECIALIZATION EXAMPLE) Projects are prepared from high impact publications relevant to the specialization. Public domain datasets are curated to prepare focused assignments illustrating how the data is processed and used to achieve similar results to the publication. Other approaches that can perform a similar function are discussed, providing a review of the methods section of the paper. Finally, full dataset is organized into a project format that can be analyzed for discovery.
- 17. 17 FALL 2018 INDIA: USA: ONLINE: Louisiana Biomedical Research Network – September 2018 Georgetown University – October 2018 August 25 – workshop in Kolkata (APT SOFTWARE AT SALT LAKE) September 6 – AIIMS (AIIMS, THEATRE ROOM ) September 15: Short Term Training in Next Generation Sequencing (Transcriptomics and Genomics) October 15: Short Term Training in Machine Learning for Biomedical Data
- 18. 18 server.t-bio.info Login: test@pine-biotech.com Password: WF(9iobE Files: bit.ly/2vNovGC
- 19. T-BioInfo is designed for processing, analysis and integration of multi-omics data. The platform is used in multiple research groups to extract meaningful insights from large multi-omics datasets. Our current effort expands to education, by enabling more people to extract meaningful, data-driven insights from omics datasets with biomedical applications. To learn more about the platform and it’s research and educational features, follow the highlighted links . T-bio.info | edu.t-bio.info | server.t-bio.info 19
- 20. Modeling Precision Medicine Machine Learning for Transcriptomics Data: Extracting Meaningful insights from high-throughput biomedical data. 20
- 21. Clinical Subtypes Molecular Subtypes 21
- 22. Diagnosis, Prognosis, Response to Treatment 22 Survival prediction Treatment Selection OncotypeDXPAM50
- 23. Daemen et al., 2013, “Modeling precision treatment of breast cancer”: an analysis of over 70 different Breast Cancer cell lines and over 90 different therapeutic agents. https://genomebiology.biomedcentral.com/articles/10.1186/gb-2013-14-10-r110 Sample 1 Sample 2 Sample 3 Sample 4 gene 1 4 3 3 7 gene 2 6 5 5 8 gene 3 6 6 6 6 gene 4 1 2 1 2 gene 5 9 10 1 5 gene 6 12 4 0 5 gene 7 1 7 9 8 gene 8 4 8 3 10 23
- 24. 24 server.t-bio.info Login: test@pine-biotech.com Password: WF(9iobE Files: https://bit.ly/2n54U0A
- 25. Files we will use in this session 25 /export-data/sciservice/data/pipelines/5629cc88e71b7bf5/upload/SRR925687_1.fq /export-data/sciservice/data/pipelines/5629cc88e71b7bf5/upload/SRR925687_2.fq /export-data/sciservice/data/pipelines/5629cc88e71b7bf5/upload/SRR925697_1.fq /export-data/sciservice/data/pipelines/5629cc88e71b7bf5/upload/SRR925697_2.fq
- 26. Part 1: RNA-Seq Processing from raw reads to a table of expression 26
- 27. RNA-Seq: overview .…TCTGAAACAATGCTTCAATCTAACTTATCATTCATTGGGA…. Genome 27 Gene A Gene B Gene C Transcr. ATranscript A Transcr. ATranscript C
- 28. 28 .…TCTGAAACAATGCTTCAATCTAACTTATCATTCATTGGGA….Gene A Gene B Gene C Transcr. ATranscript A Transcr. ATranscript C Reads RNA-Seq: overview
- 29. Preprocessing: • Adapters removal plus additional trimming • Removing PCR duplicates 29 Quantification of expression levels Mapping • Mapping on the set of known transcripts • Mapping on genome (and potential identification of novel transcripts) • Combined strategy RNA-Seq: overview
- 32. 32 Expression Table Sample Name Gene ID What is this number?
- 33. 33
- 34. Part 2: Machine Learning Data exploration and classification 34
- 36. Unsupervised analysis: PCA 36 • Explore data • Visualize Why use Principal Component Analysis? • Data Filtering • Outliers • Interpretation Considerations:
- 38. 38 Unsupervised analysis: PCA PCA 7,000 genes PCA PAM50 (35) genes Normal-like Basal Claudin-low Luminal
- 39. 39 Unsupervised analysis: Hierarchical Clustering • Identify groups • Associate sample to group Why use clustering? • Various methods • Random selection in some methods • Interpretation Considerations:
- 40. 40 Unsupervised analysis: Hierarchical Clustering
- 41. Unsupervised analysis: hierarchical clustering Dendrogram 41 2 clusters 4 clusters 8 clusters
- 42. BREAK 42 Q&A
- 43. 43 DogsCats ????? Training Set Test Set Supervised Machine Learning
- 44. 44 Step-wise Linear Discriminant Analysis (swLDA)
- 45. 45 Support Vector Machine (SVM) with Linear Kernel d d
- 46. 46 Support Vector Machine (SVM) with Linear Kernel ?
- 47. ? 47 Support Vector Machine (SVM) with Linear Kernel
- 48. BREAK 48 Q&A
- 49. Part 3: Interpretation Annotating and Interpreting Gene Expression 49
- 50. Gene annotation: ENSG to Gene Symbols plus GO 50
- 51. BREAK 51 Q&A
- 52. 52 Download and Modify R Scripts
- 53. 53 Upcoming Programs and Registration: Elia Brodsky Co-founder and CEO Pine Biotech (pine-biotech.com) bit.ly/2KG7iEj https://edu.t-bio.info/workshops/ elia@pine.bio; sahil@pine.bio