Tweet
I have been thinking about this for a while, when making some albeit simple heuristics for detecting the difference between sequencing error and heterozygosity. I noticed that traditional p-values were very sensitive to even small differences in errors if there were enough data ie H0: p<=.05 HA: p>.05, where .05 was the sequencing error, if there were enough data at a position even p=.949 would be significant. It seems that if you had enough data even the rounding errors in the CPU's would be significant. Is this a general weakness of the p-value theories in statistics, and is there any robust statistic, maybe called an r-value, that takes into account the magnitude of the difference, or is this something that can only be handled with unit based values? It seems like there should be a way to weight the statistics such that more data that is collected that it be more difficult to produce significant results.
-
-
Yeah what you're looking for is something that essentially ignores sample size since in these tests the concept of sample size isn't correctly applied anyways. Maybe something based on binning counts and comparing histograms with a bootstrap and simple Euclidean distance./* Shawn Driscoll, Gene Expression Laboratory, Pfaff
Salk Institute for Biological Studies, La Jolla, CA, USA */ -
No, sample size is key for smaller sized samples. I was thinking maybe picking the minimum difference in means I was comfortable with, then calculating the size of the smallest sample that that could be significant then scaling the size of the sample size downward as it approaches that threshold.Originally posted by sdriscoll View PostComment
-
Love this - good data definitely starts from good input, and poor input can only give relatively poor data. I particularly like the mention of Nanodrop/absorbance based methods for quantification. It's such a toss up if you'll get an accurate reading or what amounts to a randomly generated number, and a lot of library/sequencing related issues can be traced back to poor quant.07-01-2026, 11:43 AM
-
I’m not a sequencing expert. I’m a purification scientist who uses NGS to evaluate workflows my group develops. With this perspective, we think about the sample first and the NGS workflow second. The sequencer is an exceptionally honest reporter, but it can only report on what you give it, so whether you get clean, interpretable data from an NGS workflow is largely determined before you begin.
Here are nine questions we think about, in roughly the order they matter, before...
Comment