Preprocessing of the dorsolateral prefrontal cortex CMC gene expression dataset was performed largely in accordance with Fromer et al. [8]. This included filtering of lowly expressed genes (> 1 CPM in at least 50% of samples), conditional quantile normalization, and winsorization of gene counts, setting values in specific samples that deviate > 3 standard deviations from other samples to 3 standard deviation limit. Sample outliers were removed based on (1) visual inspection of the first two principal components of the full gene expression matrix and (2) interarray correlation, removing samples with correlation less than 3 standard deviations below mean for the dataset. For application of the CONFETI method, CMC dosage data imputed to the TOPMed reference panel, pruned at R-squared = 0.7 and with MAF ≥ 0.01 was also provided as input.
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.