• Document: Differential gene expression analysis using RNA-seq
  • Size: 1.57 MB
  • Uploaded: 2019-03-14 13:17:41
  • Status: Successfully converted


Some snippets from your converted document:

Differential gene expression analysis using RNA-seq Applied Bioinformatics Core, August 2017 Friederike Dündar with Luce Skrabanek & Ceyda Durmaz https://abc.med.cornell.edu/ Day 4 overview •  (brief) theoretical background for DE analysis •  DE analysis using DESeq2 •  exploring the results •  suggested: DE analysis with edgeR and/or limma DIFFERENTIAL GENE EXPRESSION Bioinformatics workflow of RNA-seq analysis Images .tif Base calling & demultiplexing Bustard/RTA/OLB, CASAVA FASTQC Raw reads .fastq Mapping STAR RSeQC Aligned reads .sam/.bam Counting HTSeq, featureCounts Descriptive Read count table .txt Normalizing plots DESeq2, edgeR Normalized read count table .Robj List of fold changes & statistical values .Robj, .txt Downstream analyses on DE genes Bioinformatics workflow of RNA-seq analysis Images .tif Base calling & demultiplexing Bustard/RTA/OLB, CASAVA FASTQC Raw reads .fastq Mapping STAR RSeQC Aligned reads .sam/.bam Counting HTSeq, featureCounts Descriptive Read count table plots .txt Normalizing DESeq2, edgeR Normalized read count table .Robj DE test & multiple testing correction DESeq2, edgeR, limma List of fold changes & statistical values .Robj, .txt Downstream analyses on DE genes Read count table DE basics 1.  Estimate magnitude of DE taking into account differences in sequencing depth, technical, and biological read count variability. logFC 2.  Estimate the significance of the difference accounting for performing thousands of tests. (adjusted) p-value 1 test per gene! H0: no difference in the read distribution between two conditions Garber et al. (2011) Nature Methods, 8(6), 469–477. doi:10.1038/nmeth.1613 s in this ESeq2 section as well. The di↵erentia model orm: Modeling read counts (DESeq) model and all the fitted steps taken mean in the software gene-specific dispersion are descr mula and descriptions in this sectionparameter as well. The di↵e (fitted towards the nearK ij ⇠ofNB(µ model ij , ↵i ) average dispersion) the form: read counts for Kij ⇠ NB(µij , ↵i ) gene i and library size expression µij = sj qij sample j factor value estimate µij = sjKqijij ⇠ NB(µij , ↵i ) moderated log-fold log2 (qij ) = xj. i µij = sj qij change for log2 (qij ) = xj. i gene i log2 (qij ) = xj. i

Recently converted files (publicly available):