RAPIDR(NIPT analysis ) report error

Question

[winter_li]

HI, when I build reference to use RAPIDR( NIPT analysis package), I met the error information as the below.

RAPIDR errors:

> ref.set  <- createReferenceSetFromCounts(counts.fname, outcomes, gcCorrect = FALSE, PCA = FALSE, method = "NCV")
Loading binned counts file
Checking every sampleID has an outcome
Error in createReferenceSetFromCounts(counts.fname, outcomes, gcCorrect = FALSE,  : 
  Requires at least 2 samples of females and 2 samples of males.

what happened ? why ?

Answer 1

[User 59]

Do you have at least 2 samples in the outcomes data.frame for each Gender?

What does head(outcomes) give you? It should look something like:

  SampleID     Dx Gender
1     1000 Normal Female
2     1001 Normal Female
3     1002 T21    Female
4     1003 Normal   Male
5     1004 Normal   Male
6     1005 Normal   Male

[winter_li]

HI, I do not know the sex information of every sample, So I assumed every sample is Male . Am I wrong ? I am the new guy to use RAPIDR, can you give some help ? Thank you very much!

[User 59]

It wont work if you don't have at least two samples of each gender. If you don't know what gender your samples are, and you have the BAM files, look for coverage on the X and Y chromosomes, you should be able to make an informed decision then. RAPIDR will not work out the genders for you.

[winter_li]

HI,Daniel, after thought, because it's unnecessary to judge the sex information of unknown example , so it's not a problem for that I do not know sex information , Am I right ?? I just wanna to judge T21,T18,T13, never caring about sex information of every unknown test example .

[User 59]

It's a problem for you because the software you're using requires you to know that information. You're trying to create a reference set aren't you? Iif you don't know what gender your references are, how are you sure what their Dx is?

[winter_li]

Dear Daniel, another guy provided me with some samples , just have Dx information , but have no sex information. So I think maybe sex information is not necessary for the testing of T21,T18,T13 in RAPIDR. I am not sure . Thank you ! And would you like to tell me your e-mail address ? thanks!

[winter_li]

HI, Do you have mask.bed file from RAPIDR ? I can not find that file from that package ....

[User 59]

No I don't, but you're right that it is not included in the CRAN package. You can use a bed file of repeatmasker regions from UCSC or other genome browser.