From Genome Analysis Wiki
Jump to: navigation, search

This is the Frequently Asked Questions page for the METAL Program

Metal checks whether the direction of effect is positive or negative when combining information across samples... What happens if I am using odds-ratios which are always positive?
This is probably one of the most common questions for METAL users. When carrying out a meta-analysis, METAL needs to know if two or more samples show the same or opposite direction of effect. Typically, METAL checks whether the value in the EFFECT column is positive or negative. If the EFFECT column includes an odds-ratio or relative risk, you actually need METAL to check whether the log-odds ratio is positive or negative. Fortunately, this is easy to do. All you have to do is to modify your EFFECTLABEL command to tell METAL it should take the log of whatever value is listed. For example, if the EFFECT column is labelled ODDS, use the syntax EFFECTLABEL log(ODDS)
Metal only displays the first 20 error messages of each kind. Can I get more details on which markers generated problems?
Yes. Use the MAXWARNINGS to increase the warning limit.
If Metal reports a p-value such as 1.23e-05, what does that mean?
A value like 1.23e-05 is shorthand for 1.23 x 10-5.