You might consider 1000genomes.
I'm using coordinates from GRCh37.p13:
f=http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/ALL.chr17.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf.gz tabix $f 17:41196312-41277500 | gzip > BRCA1.vcf.gz zcat BRCA1.vcf.gz | wc -l 2052 f=http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/ALL.chr13.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf.gz tabix $f 13:32889617-32973809 | gzip > BRCA2.vcf.gz zcat BRCA2.vcf.gz | wc -l 2286
I would recommend using the Indian Genome Variation Browser
You have the ability to search for a gene/region e.g incase of BRCA1. Once there you can select the SNPs from their phase1/2 data and get the frequency distribution amongst different Indian populations e.g for rs8176265
I'd also add ExAC which has a much larger cohort than 1000 Genomes and includes allele frequencies for south asian populations. You cannot sub divide that into Indians though.