Importantly, we could not compensate for these biases by simply increasing the p-value stringency when using multiplicatively scaled data, as many moderately significant, but highly reproducible genes such as rv3696c would be lost alongside the potential false positives with broad variances in p-values ( Figure S3C ).
← all excerpts
ARTIST: high-resolution genome-wide assessment of fitness using transposon-insertion sequencing.
1
—
—