On date unavailable, commented:

None

On 2014 Aug 05, Michael Baudis commented:

Thanks for the notes, and your usage of our resource. Regarding the other comments, you are obviously perfectly right that one has to do a critical evaluation of the data, beyond what we can provide.

But regarding the specific criticism about the DCIS data, I am not sure if I can find real issues:

• For Progenetix, by accepting only publication derived data, we rely on the peer review process running due course - there will be errors, but limited through standard operating procedures inherent to the process.
• All the 76 samples, as far as I can follow this up without producing a review of the field, had been labeled as DCIS in the papers or supporting information; so the number stands. Frequently the DX contained addl. specifications (e.g."poorly differentiated").
• Most importantly, for the analysis results it doesn't matter if the clinical diagnosis was incomplete due to needle biopsy (i.e. if the patient had an undetected infiltrating component) as long as the profiling was derived from the DCIS cells. Also, in hindsight and with some thousands of IDC available through www.progenetix.org and www.arraymap.org, the CGH profiles overall support the DCIS statements.

However, we are absolutely delighted about community support in improving our annotations! So if you have clear per sample corrections (and there will be plenty of opportunities I guess) - just help us through providing specific comments.

Best,

Michael.

On 2014 Aug 05, Swapnil Rane commented:

It is fantastic effort on part of the authors to provide such a resource. However, I noticed few problems on running a sample query. I searched for ductal carcinoma in situ cases in this database and initially I was very happy to see copy number information on 76 cases. But on looking at the 5 papers from which the query sourced the information, only two of them seemed valid sources. The remaining three papers according to me are not valid, as they either do not report the CNA in DCIS separately from the coexisting IDC or the technique itself by which samples were obtained(FNAC) for genetic analysis did not differentiate between DCIS and IDC. Even if I took into account only those two papers which exclusively report on DCIS and the pure DCIS cases from other three papers, the total number of cases does not exceed 43.... How did they reach a number of 76?.. Atleast, the database should annotate that the data is derived from mixed cohort of pure DCIS and DCIS with coexiting IDC. While I have not run any other query on the resource and it is possible that this glitch is only seen for this particular query, but it has made me sceptical about the reliability of the information provided by any such resource. It seems we cannot take any information at face value, no matter how reliable the source might seem to be!!

On 2014 Aug 05, Swapnil Rane commented:

It is fantastic effort on part of the authors to provide such a resource. However, I noticed few problems on running a sample query. I searched for ductal carcinoma in situ cases in this database and initially I was very happy to see copy number information on 76 cases. But on looking at the 5 papers from which the query sourced the information, only two of them seemed valid sources. The remaining three papers according to me are not valid, as they either do not report the CNA in DCIS separately from the coexisting IDC or the technique itself by which samples were obtained(FNAC) for genetic analysis did not differentiate between DCIS and IDC. Even if I took into account only those two papers which exclusively report on DCIS and the pure DCIS cases from other three papers, the total number of cases does not exceed 43.... How did they reach a number of 76?.. Atleast, the database should annotate that the data is derived from mixed cohort of pure DCIS and DCIS with coexiting IDC. While I have not run any other query on the resource and it is possible that this glitch is only seen for this particular query, but it has made me sceptical about the reliability of the information provided by any such resource. It seems we cannot take any information at face value, no matter how reliable the source might seem to be!!