Philosophy of Data Science | Jingyi Jessica Li | Advancing Statistical Genomics

<p>Jingyi Jessica Li | Advancing Statistical Genomics</p> <p>Watch it on….    <a href='https://www.youtube.com/channel/UCkEz2tDR5K6AjlKw-JrV57w/videos'>YouTube</a>       <a href='https://podofasclepius.podbean.com/'>Podbean</a></p> <p>Jingyi Jessica Li (UCLA) describes common statistical pitfalls in genomic data analysis & the statistical reasoning required to correct these mistakes.</p> <p>Common themes throughout include:</p> <ul><li>Hypothesis-driven science & critical scientific reasoning over data</li> <li>p-values and non-sensical null hypotheses/distributions</li> <li>the value of appearing statistically rigorous</li> <li>researchers cutting intellectual corners & digging themselves into local minima</li> </ul> <p> </p> <p>Episode Topics</p> <p>0:00 A major advancement in genomic data leads to new statistical techniques</p> <p>2:15 Hypothesis-driven science & hypothesis-free data analysis</p> <p>2:55 A ChIP Seq Example</p> <p>8:00 Misformulation of sampling variability</p> <p>16:55 A false analogy: the permutation test</p> <p>19:03 Losing my p-value religion: the value of statistical packaging</p> <p>24:30 The Clipper Framework for false discovery rate control</p> <p>31:50 Non-parametric developments</p> <p>37:55 Inferred covariates</p> <p>46:00 PseudotimeDE: inferences of differential gene expression along cell pseudotime</p> <p>47:10 Selective inference</p> <p>49:25 What biological/physiological data will be incorporated in the future?</p> <p>52:30 Statistics, computer science, data science, ML, biology</p> <p>57:05 Machine learning and prediction</p> <p>1:01:30 Sophisticated models vs sophisticated research</p> <p>1:07:45 Peer review in science</p> <p>1:13:05 Hypothesis-driven science vs cutting intellectual corners</p> <p>1:18:12 What topic should the statistics community debate?</p>