Using singular value decomposition to alleviate batch effects in RNA sequencing data — ASN Events
  1. Schedule
  2. Poster Session 2
  3. Using singular value decomposition to alleviate batch effects in RNA sequencing data

Using singular value decomposition to alleviate batch effects in RNA sequencing data (#253)

Alexandra L Garnham 1 2 , Hannah K Vanyai 3 , Tim Thomas 1 2 , Anne Voss 1 2 , Gordon K Smyth 1 4
  1. The Walter and Eliza Hall Institute of Medical Research, Melbourne, Victoria, Australia
  2. Department of Medical Biology, The University of Melbourne, Melbourne, Victoria, Australia
  3. The Harry Perkins Institute of Medical Research, Perth, Western Australia, Australia
  4. Department of Mathematics and Statistics, The University of Melbourne, Melbourne, Victoria, Australia

Batch effects can be deleterious to an RNA sequencing analysis. They have the potential to mask true results, give false positives if not identified and appropriately corrected, leading to an incorrect biological interpretation of the results. However, not all batch effects are easily discovered or rectified. One approach to correct unknown effects is to apply a singular value decomposition (SVD). By applying the SVD method to the data, we can then calculate one or more surrogate variables which adjust for these unknown factors in the analysis and allow genuine results to be identified. We demonstrate this method through an RNA sequencing analysis which investigates the gene and pathway targets of the histone acetyltransferase gene Moz in mouse palate development.

#LorneGenome