Good evening one and all...
Recently I have submitted a original research work to Journal of Herbal Medicine. In that I have received a query like " The authors have made almost all requested changes. Yet, the biggest issue, the interpretation of the GC-MS results, is still stand.
None of the halogenated acetonitriles is known to occur as a natural product. Thus, Chloromethyl Cyanide is very unlikely to be present in exanimated extract. Also, nitrosoguanidine poses carcinogenic properties and usually is used to induce cancer in experimental animal models. Deeper insight into results obtained by GC/MS analysis and following careful interpretation is necessary. Overall complete GC/MS analysis is not adequate for ethanolic extracts and I propose either serious revision of this experimental part either complete deletion of GC/MS from manuscript"
From my side response is GC-MS is only qualitative analysis and whatever compounds obtained only probability provided by NIST library based on nearby molecular weight compounds and not confirmed compounds. Further I needed to separate and confirm with spectral data.
In this regard, I needed help how can I defend for above mentioned query for my publication since myself newly working on GC-MS and its data interpretations...
Thank you.
From my side response is GC-MS is only qualitative analysis and whatever compounds obtained only probability provided by NIST library based on nearby molecular weight compounds and not confirmed compounds. Further I needed to separate and confirm with spectral data.
From what I can understand from your question the issue is this, once you obtain a list of suspect compounds based on a library search be it NIST or any other you next need to look at the suggestions and consider the likelihood that it is correct. One of the factors in screening is to add in a bit of probability based on the matrix you are studying this cannot be done by search engines. Some engines are even designed to provide the most likely from literature without considering the matrix. I use a scale of 1-5 at this stage and generate a brief synopsis of each compound suggested. In the main text I only consider the most likely based on the matrix, but reserve the rest for SI with heavy annotation of how likely/unlikely the prediction is.
Dear Dr Mahesh Kancherla,
I think what is rightly noted here is the adoption of the library search as a result without evaluating these results sufficiently.
Each library search shows a hit list depending on a calculated match of the measured spectrum to a library spectrum. However, the evaluation of this hit is always up to the user. In GC-MS analyses I have often observed hits with a match > 90%, which were clearly wrong and in return correct results with just 9%.
As Garry Codling has already explained, the matrix plays a significant role. So does the probability of the analytes supposedly found in the sample. This has also been noted by Garry Codling.
You should only use database results in your publication that you have confirmed.
Maybe you check your results again in the publication under this aspect.
Best regards
Joachim Horst
- Badges
- Science topic