On 2016 Sep 19, Amit Dutt commented:

We thank the reader for the general appreciation and interest expressed in the TMC-SNPdb initiative. We agree that using tumor adjacent normal samples from cancer patients indeed has limitations to the effect as described in an unambiguous way in our article. In addition to our study, the Exome Aggregation Consortium (Lek M, 2016) –involving 7,601 normal blood samples derived from cancer patients of 60,706 normal samples studied— has similarly described the limitation in using such normal samples. Of note, the TMC-SNPdb is a pilot initiative project. As it evolves with the inclusion of additional normal samples, we do anticipate further refinements over the subsequent release/ versions of the database.<br> Unfortunately, the suggested comparison between the TMC-SNPdb with Sudmant et al. 2015 study cannot be made as Sudmant et al. (doi:10.1038/nature15394) performed a low pass whole genome sequencing at ~8x coverage to describe “structural alterations". Such low pass coverage studies are not ideally suited for variant analysis. However, in a separate study, the 1000 Genomes project consortium (1000 Genomes Project Consortium., 2015) has described whole exome sequence data of samples including data from >400 "normal" people of South-East Asian/Indian ethnicity at a mean high coverage of ~75x to exhaustively enlist the SNPs present in the population. Our study describing the TMC-SNPdb does compare and deplete the SNPs reported from this study to develop a unique set of yet undescribed SNPs specific to the Indian population, in an exclusive manner.

On 2016 Aug 31, Chandan Kumar commented:

This seems like a welcome initiative to generate a SNP database for Indian populations. However as noted in the discussion, use of tumor adjacent normal samples derived from cancer patients to generate a reference germline database is problematic. Interestingly, the 1000 genome project (Sudmant et al, An integrated map of structural variation in 2,504 human genomes,Nature 526,75–81,(01 October 2015), doi:10.1038/nature15394) includes data from >400 "normal" people of South-East Asian/Indian ethnicity. May be useful to compare the two datasets.

On 2016 Aug 31, Chandan Kumar commented:

This seems like a welcome initiative to generate a SNP database for Indian populations. However as noted in the discussion, use of tumor adjacent normal samples derived from cancer patients to generate a reference germline database is problematic. Interestingly, the 1000 genome project (Sudmant et al, An integrated map of structural variation in 2,504 human genomes,Nature 526,75–81,(01 October 2015), doi:10.1038/nature15394) includes data from >400 "normal" people of South-East Asian/Indian ethnicity. May be useful to compare the two datasets.