An R package that allows the user to estimate the relative cell type balance in brain cortical samples using microarray or RNAseq data as presented in the manuscript "INFERENCE OF CELL-TYPE COMPOSITION FROM HUMAN BRAIN TRANSCRIPTOMIC DATASETS ILLUMINATES THE EFFECTS OF AGE, MANNER OF DEATH, DISSECTION, AND PSYCHIATRIC DIAGNOSIS" (http://biorxiv.org/content/early/2016/11/25/089391.full.pdf+html) by *Megan Hastings Hagenauer, Ph.D.1, Jun Z. Li, Ph.D.2, David M. Walsh, Psy.D.3, Marquis P. Vawter, Ph.D.3 Robert C. Thompson, Ph.D.1, Cortney A. Turner, Ph.D.1, William E. Bunney, M.D.3, Richard M. Myers, Ph.D.4, Jack D. Barchas, M.D.5, Alan F. Schatzberg, M.D.6, Stanley J. Watson, M.D., Ph.D.1, Huda Akil, Ph.D.1 1Mol. Behavioral Neurosci. Inst., Univ. of Michigan, Ann Arbor, MI, USA; 2Genet., Univ. of Michigan, Ann Arbor, MI, USA; 3Univ. of California, Irvine, CA; 4HudsonAlpha Inst. for Biotech., Huntsville, AL, USA; 5Stanford, Palo Alto, CA, 6Cornell, New York, NY, USA
Other
7
stars
4
forks
source link
Add automatic check for low marker numbers, add alternative calculations #6
Isabelle found that using BrainInABlender on older microarray data (early 2000's Affy) produced unstable results because the number of publication-specific marker genes for some cell types was often <10 or even <5. We decided that it would be best to add an automatic check for low marker number, add an error message for the publication specific indices, and then add an alternative calculation that skips averaging by publication before averaging by cell type.
Isabelle found that using BrainInABlender on older microarray data (early 2000's Affy) produced unstable results because the number of publication-specific marker genes for some cell types was often <10 or even <5. We decided that it would be best to add an automatic check for low marker number, add an error message for the publication specific indices, and then add an alternative calculation that skips averaging by publication before averaging by cell type.