Title: a Pipeline to Extract Drug-adverse Event Pairs from Multiple Data Sources

I read with interest this paper, which is aimed at describing a pipeline for drug safety signal detection. The research topic is highly relevant. As follows, some major and minor criticisms: Major revisions 1. Authors should pay carefully attention to the terminology being used. In particular, this study cannot in no way identify adverse drug reaction (implying causality assessment) but rather adverse event (based only on statistical association). I suggest reconsidering the tile and the objective of the pipeline. Moreover, correct definition of safety signal should be taken into account. Signal concerns a new or incompletely documented drug-adverse event association. Thus, by definition, a known-adverse event association should not be considered as signal if found through the pipeline. 2. Introduction: In general, introduction should be more concise. On the other hand, authors should describe some international initiatives which have been exploring electronic medical records (in comparison to spontaneous reporting system) for drug safety signal detection: a) EU-ADR (Coloma PM, et al. Based on all these ongoing initiatives, I suggest revising the statement " EHR are either not easily accessible due to privacy concerns or are very complex to understand " 3. Objective of the study should be more clearly described 4. Methods: what is the rationale for choosing the study drugs a part from being cited in other papers? Are then results generalizable to other drugs belonging to different drug classes? Authors should comment on it. In addition, the types of medication should be revised: carbamazepine is antiepileptic drug and not antipsychotic either antidepressant;. In would distinguish between antidepressants (trazodone, and paroxetine) and antipsychotics (olanzapine and ziprasidone; why not including any old typical antipsychotics which are in the market since longer period?) 5. Results: why results has been reported only for bupropion? Authors should describe results for all the study drugs trying to use visualization which may summarize results. The disagreement between different sources (as well as additional value of each source) should be better analysed and discussed. In general results should be described using more quantitative and measurable parameters 6. Table 1. Some important safety concern are not listed: a) stroke and all-cause mortality for olanzapine; b) sexual disorders for paroxetine. The list of safety concerns for each drug should be revised by experts in pharmacoviglance. Why no know AE has been reported for orphine? Minor revisions

> Thank you for your recent submission to BMC Medical Informatics and Decision Making.Before we can proceed with peer review we will need you to make some changes to the formatting of > your manuscript.We strongly urge you to make these changes promptly, as we cannot start the peer review process until we have received a version containing the changes.>> I note that you have not provided full details of your pipeline.Our editorial policies relating to manuscripts which present new software tools or services are as follows: > 1.The software application, algorithm, code or tool described in the manuscript is available for testing by reviewers in a way that preserves their anonymity.> 2. If published, the software application, algorithm, code or tool will be readily available to any scientist wishing to use it for non-commercial purposes, without restrictions such > as the need for a material transfer agreement.> 3. BioMed Central recommends, but does not require, that the source code of the software application, algorithm, code or tool should be made available under a suitable open-source > license that will entitle other researchers to further develop and extend it if they wish to do so.Typically, an archive of the source code of the current version of the software > should be included with the submitted manuscript as a supplementary file.As it is likely that the software will continue to be developed following publication, the manuscript should > also include a link to the home page for the software project.For open source projects, we recommend that authors host their project with a recognized open-source repository such as > bioinformatics.orghttp://bioinformatics.org/or sourceforge.nethttp://sourceforge.net.> Please make your code available as e.g. an additional file, in compliance with point 1.If this is not possible for any reason, please provide us with details of why this is the > case." The pipeline described in the manuscript, used to extract the adverse event pairs from traditional structured adverse event databases and non-traditional sources such as usercomments from health-related websites and MEDLINE abstracts, is a semi-automated one.Some components of the pipeline involve manual tasks.For the core of the pipeline, which includes the NLP modules that involve NER and concept association map building, an application called TPX has been modified and used.The modified version used in this pipeline is available as a web tool and can be accessed at: http://www.atc.tcs.com/tm/medra/login.jsp .Users can register on the application and test it.
>> Authors' information: Please place the Authors' Contributions section after Competing interests.Please check the instructions for authors on the journal website for the correct > format to use for Authors' Contributions.The Author's contributions are as follows: All the authors participated in the design of the study.SY and SVG were involved in the development of the various components of the software.SY and AR conceptualized the study, performed the statistical analysis and drafted the manuscript.RS reviewed the study, helped to draft the manuscript and gave final approval of the version to be published.All authors read and approved the final manuscript.

>> Competing interests:
> Manuscripts should include a ?Competing interests?section.This should be placed after the Conclusions/Abbreviations. Please consider the following questions and include a > declaration of competing interests in your manuscript: > Financial competing interests > ?In the past five years have you received reimbursements, fees, funding, or salary from an organization that may in any way gain or lose financially from the publication of this > manuscript, either now or in the future?Is such an organization financing this manuscript (including the article-processing TCS Innovations Labs Hyderabad June 28, 2013 charge)?If so, please specify.> ?Do you hold any stocks or shares in an organization that may in any way gain or lose financially from the publication of this manuscript, either now or in the future?If so, please > specify.> ?Do you have any other financial competing interests?If so, please specify.> Non-financial competing interests > ?Are there any non-financial competing interests (political, personal, religious, ideological, academic, intellectual, commercial or any other) to declare in relation to this > manuscript?If so, please specify.
All the authors are salaried employees of Tata Consultancy Services Ltd., where this work was done and the organization finances the article-processing charge for this manuscript.The authors have no other competing interests.
> ?Do you hold or are you currently applying for any patents relating to the content of the manuscript?Have you received reimbursements, fees, funding, or salary from an organization that holds or has applied for patents relating to the content of the manuscript?If so, please specify.
TPX-PV uses various dictionaries such as chemical & drugs, disease, symptoms curated automatically using a dictionary curation pipeline for identifying entities from the unstructured text (viz., Blogs, PubMed).A patent has been filed on this dictionary curation pipeline.The details of the patent are mentioned below : Patent Name : "AUTOMATED DICTIONARY CREATION FOR SCIENTIFIC TERMS" Patent Application Numbers : 13752620, EP13153222 and 298/MUM/2012 in US, Europe and India respectively.
> For more information please visit the instructions for authors: <http://www.biomedcentral.com/bmcmedinformdecismak/authors/instructions/researcharticle#formatting-competing> > We would be grateful if you could address the comments in a revised manuscript and provide a cover letter giving a point-bypoint response to the concerns.--------------------------------> When you have revised your manuscript in light of the reviewer's comments and made any required changes to the format of your paper, please upload the revised version by following > these instructions: > 1. Go to http://www.biomedcentral.com/manuscript/login/man.asp?txt_nav=man&txt_man_id=1822012469101544 and log on with your email address and password.> 2. With the 'Manuscript details' tab, please update the title, abstract and author details if they have changed since the Tata Consultancy Services Ltd.
To Extract Drug-Adverse Event Pairs From Multiple Data Sources > SriJyothsna Yeleswarapu, Aditya Rao, Thomas Joseph, Vangala G Saipradeep and Rajgopal Srinivasan > Dear Mr Rao, previous version.It is very important that all changes are > updated on this page, as well as in the manuscript file as the information on this page will be used in PubMed and on BioMed Central if your manuscript is accepted for publication.> 3.With the 'Cover letter' tab, please provide a covering letter with a point-by-point description of the changes made.> 4. With the 'Upload files' tab, please upload the revised version of the manuscript and press 'Submit new version'.Please wait for the confirmation page to appear -this may take a > few moments.
TCS Innovations Labs HyderabadJune 28, 2013 > Please also ensure that your revised manuscript conforms to the journal style (http://www.biomedcentral.com/info/ifora/medicine_journals).It is important that your files are > correctly formatted.>Welookforward to receiving your revised manuscript by 30 June 2013.If you imagine that it will take longer to prepare please give us some estimate of when we can expect it.>Youshould upload your cover letter and revised manuscript through http://www.biomedcentral.com/manuscript/login/man.asp?txt_nav=man&txt_man_id=1822012469101544.You will find more > detailed instructions at the base of this email.>Please don't hesitate to contact me if you have any problems or questions regarding your manuscript.