I make all my papers available through preprint servers when they are ready, and no longer see any reason to send them to journals, even open access ones. My current research interests involve applying machine learning to genomics datasets. Bill Simmons and Kevin Wildes are known for a podcast in which they pitch half-baked ideas. Some ideas are good, some not so much. Similar to this podcast, in addition to discussing my complete manuscripts I will pitch ideas or provide initial steps in analyses that could be completed by someone with more experience and interest in the field. If you like what you see you can request a specific analysis for your lab (see Contact for more details).
"Half-baked" Analysis: GRIMMEST: A method to test the validity of test statistics Posted December 27th, 2016 The GRIM test checks means and the GRIMMER test checks standard deviations. In this post I describe how granularity testing can be used to detect inconsistencies in test statistics. Continue reading
Publication: The GRIMMER test Posted September 11th, 2016 With this discovery I am one of the founding members of the statistical subfield of granularity testing. In this post I describe my contributions to the field.
Publication: PrePubMed: A PubMed for preprints Posted June 10th, 2016 Posts about this paper can be found here
Publication: OncoLnc: Linking TCGA survival data to mRNAs, miRNAs, and lncRNAs Posted February 17th, 2016 Article is available at https://peerj.com/articles/cs-67/ Posts about this paper can be found here
Publication: Pan-cancer analysis of prognostic genes Posted November 9th, 2015 The paper is available at https://peerj.com/articles/1499/ Posts about this paper can be found here
"Half-baked" Analysis: Hepcidin's role in atherosclerosis Posted November 1st, 2015 This is a grant that I wrote as a training exercise. If you are interested in hemochromatosis, iron regulation, atherosclerosis, or interesting applications of Brainbow, check out this post. Continue reading
"Half-baked" Analysis: Using MAF files to find neo-antigens Posted August 12th, 2015 Immunotherapy is an exciting new area of research in cancer biology, and identifying neo-antigens that can illicit an immune response is an essential component. With the advent of whole genome sequencing, it is possible to find all of the mutations in a tumor, and these mutations are typically listed in a mutation annotation format file. Using these files to find neo-antigens requires some technical expertise and knowledge of the immune system. Continue reading
"Half-baked" Analysis: Using ProteomicsDB to analyze the human proteome Posted August 10th, 2015 ProteomicsDB is a new resource that is likely underutilized due to the technical expertise required to use an API. This mass-spec database contains a wealth of information, including different kinds of protein modifications. Continue reading