search

gagneurlab / FRASER

FRASER - Find RAre Splicing Events in RNA-seq
MIT License
36 stars 20 forks source link

Question regarding the sample size of FRASER #21

Closed lyj95618 closed 3 years ago

lyj95618 commented 3 years ago

Hello. I am new to bioinformatics and curious how to use FRASER to look for aberrant splicing events in a single sample (If this is possible)? I know the recommended sample size is 30. Lets say if I run 1 patient sample and 30 normal samples from GTEx together, will this sufficient enough to look for aberrant splicing events in my sample? Any help will be much appreciated.

c-mertes commented 3 years ago

Dear @lyj95618, this is related to #10

FRASER needs some data to fit its model. Having just one sample will make it hard. Merging it with GTEx sample can help but we would recommend having at least 10 samples to merge with GTEx. The problem with 1 sample is that any technical bias will pop up as an outlier event, as the model can not model the covariates between the GTEx samples and your single sample.

For more details on how to merge external datasets, you can have a look at our new protocol https://www.nature.com/articles/s41596-020-00462-5

lyj95618 commented 3 years ago

Thank you very much for your detailed explanation!