X team 2 - presentation
Download as PPTX, PDF1 like539 views
The document discusses the challenges and solutions regarding high-dimensional biological data analysis in biomedical research, focusing on the need for effective tools, data accessibility, and training. It highlights the importance of integrating various data types and addressing issues like data fragmentation and quality. Additionally, it promotes interdisciplinary collaboration and the creation of resources for data science methodologies.
X team 2 - presentation
- 1. X-team #2 High Dimensional Biological Butterflies Data Science Workshop 2015
- 2. What do we have in common?
- 3. High-dimensional biological data ● High-throughput genotyping and phenotyping ● Finding biological meaning in big data with high N and/or P
- 4. The ability to harvest the wealth of information contained in biomedical Big Data will advance our understanding of human health and disease; however, lack of appropriate tools, poor data accessibility, and insufficient training, are major impediments to rapid translational impact. -NIH BD2K
- 5. Data integration ● Data fragmentation o individual vs population o multiple -omics o multiple sources ● Discovery and prediction o genome and functional annotation Statistical learning methods ● Data quality ○ hidden sources of variability ○ limitations of short read sequencing Data annotation Genome assembly/error correction Problem Solution
- 6. Success Stories Domain Science Data Science Methods Metabolic pathway - Ingenuity Pathway Analysis (http://www.ingenuity.com/products/ipa) Genomic data - Quality Control - FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) - EasyQC for genome-wide association meta-analyses (http://www.nature.com/nprot/journal/v9/n5/full/nprot.2014.071.html) - Batch effect - PEER (http://www.ncbi.nlm.nih.gov/pubmed/22343431) - SVA (http://www.ncbi.nlm.nih.gov/pubmed/22257669) - scLVM (Buettner et al., 2015) - Data storage and sharing - NCBI (http://www.ncbi.nlm.nih.gov) - GitHub (https://github.com) - UCSC genome browser (http://genome.ucsc.edu/) - Gene annotation - Gene Ontology (http://geneontology.org/page/documentation) Proteomics - Protein Data Bank (PDB) (http://www.rcsb.org/pdb/home/home.do) Disease Survivability - WEKA (Mark Hall, Eibe Frank, Geoffrey Holmes, Bernhard Pfahringer, Peter Reutemann, Ian H. Witten (2009); The WEKA Data Mining Software: An Update; SIGKDD Explorations, Volume 11, Issue 1.)
- 7. Same data, different interpretation Gilad & Mizrahi-Man 2015 F1000Research, 4:121
- 9. Going Forward ● Create and maintain a HowTo website for Data Science computational tools and methods. http://data-science-for-biologists.wikia.com/wiki/Data_Science_for_Biologists_Wikia ● Collaborate via Github
- 10. Thanks!
Editor's Notes
- #3: Half are domain scientists and half are more computationally inclined. Made this word cloud from out notes. Data. Comp bio. Disease. Genetics. Integrative anlyses.. Disease spread. Social environment and epigenetics. Data privacy, data sharing, and computational genetics. Genetic and Proteomics and statistical tool to understand disease and cancer or individual phenotypic variation Tool development. RNAseq technology and applications tools for data reduction and variable selection.
- #4: S
- #7: predicting disease survivability for breast cancer patients Famous example: Potential flaws in genomics paper scrutinized on Twitter:http://www.nature.com/news/potential-flaws-in-genomics-paper-scrutinized-on-twitter-1.17591