Case-control study power analysis help

[ciphered lore]

Went through Google results but didn't find a straightforward way to calculate power for SNP and CNV detection. What's the best approach for rare versus common variants? Hoping someone can share insights.

 

[orphan 99]

For common SNPs, usually standard power calculators will work, requiring details like allele frequency, effect size, disease prevalence, desired statistical power, and significance level.

as for rare changes, figuring out the power is trickier. You often need to group these rare changes together. Power in this case depends on how many rare changes are in that specific area, their combined effect and how many people you need to study to see that combined effect.

Have you looked at any specific software tools for power analysis in genetics like GCTA, CaTS? this might help you with your search as well! good luck!

 

[ciphered lore]

For common SNPs, usually standard power calculators will work, requiring details like allele frequency, effect size, disease prevalence, desired statistical power, and significance level.

as for rare changes, figuring out the power is trickier. You often need to group these rare changes together. Power in this case depends on how many rare changes are in that specific area, their combined effect and how many people you need to study to see that combined effect.

Have you looked at any specific software tools for power analysis in genetics like GCTA, CaTS? this might help you with your search as well! good luck!

Wow, thank you! That breakdown is a real lightbulb moment for me. The software suggestions are the missing piece of the puzzle I was looking for. I will start investigating GCTA and CaTS right away.

 

[Low_Local26]

For common SNPs, usually standard power calculators will work, requiring details like allele frequency, effect size, disease prevalence, desired statistical power, and significance level.

as for rare changes, figuring out the power is trickier. You often need to group these rare changes together. Power in this case depends on how many rare changes are in that specific area, their combined effect and how many people you need to study to see that combined effect.

Have you looked at any specific software tools for power analysis in genetics like GCTA, CaTS? this might help you with your search as well! good luck!

Thanks for sharing! I'm just starting out so just would like to ask, how do tools like GCTA or CaTS actually handle rare variants? Do they let you group them easily, or do you need to prep the data a certain way first?

 

[ciphered lore]

Thanks for sharing! I'm just starting out so just would like to ask, how do tools like GCTA or CaTS actually handle rare variants? Do they let you group them easily, or do you need to prep the data a certain way first?

Hey, good question. I'm still getting my bearings too, but from what I've seen so far, it looks like you do need to prep the data a bit before feeding it into tools like GCTA or CaTS. They don't automatically group rare variants for you, you'd probably need to define those regions or sets ahead of time, maybe based on gene boundaries or functional annotations.

I'm planning to test out GCTA next week once I finish cleaning up my sample list. If I hit any snags with rare variant handling, I'll circle back here and share what I find. If you end up trying it 1st, I'd love to hear how it goes on your end, too.