Referee 1's Review (David Goldgar):

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

Review of the original submission:

This paper presents the first analysis of BRCA2 mutations in the Malaysian population and is thus of interest. I have a few comments/suggestions.

1. In the abstract,when mentioning "early-onset disease", please specify the age criterion.

2. In the abstract, under conclusions, the statement that there is a higher frequency of BRCA2 than BRCA1 mutations is over-stated, given that it is 2 vs. 1 out of 37. Unless there is a statistically significant difference, probably best not to comment on this. Also the figurs of 2.7% and 5.4% are mentioned twice in this paragraph.

3. In the introduction, two additional studies carried out in this population are not referenced: Patmasiriwat P et al. 2002 and Zhi et al. 2002.

4. In the results section, the statement that "the age-standardised incidence is higher.." should be referenced.

5. In tables 1 and 2, I don't see the need to report all the identified common polymorphisms. Perhaps just state that the known polymorphisms are found in this population as well.

6. I would prefer the authors use the term "VUS" for variant of uncertain significance, rather than "UCV".

7. In the section on BRCA1 Variants, it should be mentioned that the variant R726S is not in one of the recognized functional domains and thus is unlikely to be pathogenic. Moreover, in contrast to the authors statement, it is not conserved and in fact the Serine residue is present within the alignment.

8. The authors should state whether any of the synonymous variants are potential splice variants (e.g., in the last nucleotide of the exon).

9. In the conclusions, the discussion of the relative frequency of BRCA2 to BRCA1 mutations should be toned down (see point 2).

Review of the first revised manuscript:
The authors have addressed most of the concerns raised in my previous review. However, I note that the abstract still contains a sentence highlighting an increased frequency of BRCA2 variants compared to BRCA1. This is clearly not the case. If we focus on the deleterious and VUS there are 5 BRCA2 vs. 3 BRCA1. This is clearly not different from 50-50. Of course one would expect more polymorphisms/neutral rare variants in BRCA2 given it is almost twice as large as BRCA1. As I indicated before, any such comparison should be removed. I also think the polymorphisms part of tables 1 and 2 should perhaps be put in a separate table, as most of the columns are not really relevant (e.g., BIC entries. It would be nice to provide a comparison of the allele frequencies of these polymorphisms in this group with those in other populations. Also the 'AA Change' entry for the IVS3 mutation should be blank or NA or something. Lastly, all mutations should be in HGVS notation with cDNA sequence and protein sequence. This is now becoming a requirement for all publications reporting sequence variant data.