sumstats.txt.gz file 

Hello!
The sumstats file consists of the original beta and p value of the GWAS for each SNP? 
If yes, should we take the negative of the beta in case the effect allele in the one study is the reference allele of the other?  
Should this be done also in the input files of ldsc_preprocess?
Thanks!