GCTA conditional analysis

1. --cojo-cond and --cojo-slct are two distinct analyses.
2. --cojo-cond runs the conditional analysis once given the list of SNPs provided.
3. --cojo-slct runs the conditional analyses iteratively each time picking up the top SNP. Results presented in Yang et al. 2012 Nat Genet Table 1 were from this analysis.
4. For --cojo-cond, the list of SNPs provided should not have a co-linearity issue, otherwise the model is not solvable.
5. I guess what you need is conditional p-values of all other SNPs given a list of top SNPs.
When you can run
a) /.gcta--bfile gonlchrbp
--chr 9
--cojo-file metaconditional.txt
--extract conditional1000inGONL.txt
--cojo-slct
--out test

b) take the SNPs from output file *.jma, then run
/.gcta--bfile gonlchrbp
--chr 9
--cojo-file metaconditional.txt
--cojo-cond SNPs_taken_from_analysis_above
--out testconditional9

This procedure gets rid of the co-linearity problem in your list of given SNPs.

However, I would still suggest you run --cojo-slct. It pick up everything conditionally significant.
/.gcta--bfile gonlchrbp
--chr 9
--cojo-file metaconditional.txt
--cojo-slct
--out test_slct