Problem while replicating methylation predicting paper, looking for explanation
1
0
Entering edit mode
13 days ago
Junk • 0

I'm trying to replicate this paper (https://pathsocjournals.onlinelibrary.wiley.com/doi/10.1002/path.5845) and I don't understand this bit from the "Data Preprocessing" section on page 4:

Beta values whose corre-sponding intensities on probe-level failed a z-test with a P value threshold of 0.01 against background signal were masked.

Can someone please explain what this means? I'm using R, mainly minfi.

bioconductor methylation machine-learning r • 313 views
ADD COMMENT
2
Entering edit mode
13 days ago
cmdcolin ★ 4.0k

here is the output from google gemini when I ask it to translate the sentence to more plain english:

"This sentence is explaining a data quality control step in the analysis of methylation chip data. Here's the breakdown in simpler terms:

  • Methylation chips: These are tools used to measure how much DNA is chemically modified (methylated) at specific spots in the genome.
  • Beta values: These are numbers that represent the level of methylation at each spot on the chip.
  • Probe-level intensities: These are measurements of the signal strength from each spot on the chip.
  • Z-test: This is a statistical test that compares the probe-level intensities to a background signal. (me note: Z tests are based on normal distribution/population standard deviations)
  • P-value: This is a number that indicates the statistical significance of the difference between the probe-level intensities and the background signal.
  • Masked: This means that the beta values were removed from the analysis because they did not pass the quality control check.

In essence, the researchers removed any beta values where the corresponding signal from the chip was too weak or noisy compared to the background noise. "

to dive deeper into these topics, you can check out this intro methylation analysis https://life-epigenetics-methylprep.readthedocs-hosted.com/en/latest/docs/introduction/introduction.html

here is also a fairly accessible paper about the choice of p-values for filtering probes here https://pmc.ncbi.nlm.nih.gov/articles/PMC6346546/

ADD COMMENT

Login before adding your answer.

Traffic: 1950 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6