I blogged recently about the Publishing Process and when the publishers get it right. That was my blessing of their work. This posting has a slightly different flavor…it’s about my concerns regarding the editorial and peer-review process. This posting is not meant to be generally applicable to all journals nor all editors. It is simply a review of recent experiences. It’s NOT about ChemSpider per se, just a general rant about my experiences of publishing and reviewing science, something we will be doing in the future about our experiences with ChemSpider.
In my line of work I get to work with some excellent scientists. Similarly, a number of the people I have developed respect for now make up the ChemSpider Advisory Group. As a result of scientific collaborations I author/co-author about 8 peer-reviewed articles a year. I also review about 6 articles a year for various journals. Of late the processes involved have me concerned.
Some of my recent experiences having articles I am involved with reviewed are discussed below.
1) One article had multiple “Publish As-is” with minor grammatical corrections. One reviewer clearly didn’t have an understanding of the science and got stuck on “it’s too long”. After a long discussion with the editor, despite the multiple “Publish as-is” we were left with no choice…reduce the length of the article or retract. We reduced the length of the article. I suspect that there are page limits being put on editors to the disserviceÂ of good science.
2) A recent submission to one of the most popular journals in the world was sat on for many months. This work resulted from work with a pharmaceutical company and the submission therefore was from a chemistry software company and a pharmaceutical company.The manner in which this work was treated suggests unfair bias to non-academic submissions as if the fact that the work was included into a commercial product was problematic. As a result of the treatment received we have committed to not return to that journal to publish.
3) I was asked to review a publication including a comparison of performance of one of our prediction algorithms. A table contained a comparison of numerous algorithms including ours. The article in question highlighted that the authors algorithm outperformed any of the other algorithms under comparison. A close examination showed that for our algorithm four numbers had been mixed up in the table and arranging them properly showed that actually our algorithm had the best performance. This feedback was given to the editors. The paper was then published with MORE errors than originally identified and our algorithm was way down the list. A request to the editor for an Erratum or publication of a rebuttal by us was turned down…specifically because they did not feel it was appropriate to get in the middle of such a comparison with a “commercial organization.” It took two years but the original author was kind enough to publish a retraction with us but by that time the damage was done.
My recent experiences with articles I have reviewed include two in particular (no names or topics mentioned)
1) I was asked to review an article that was categorically “bad science”. I strongly suggested that the article be retracted..not rewritten. The science behind the article was simply wrong. There was nothing to rescue. The article was withdrawn but published very shortly after with 70% of the content intact in an Open Access journal (no title mentioned). It remains bad, but published, science.
2) I was asked to review an article by an ACS journal editor. This was another one comparing the performance of a series of algorithms, one of ours included. The work made our algorithms look terrible. It did the same for many others. Some of the data used to build the correlation were extracted from other publications. Some of the data were not even what they were correlating. A close examination of the data showed that when bad data were removed and the remaining data were treated appropriately that our algorithms gave excellent performance. I provided detailed feedback to the editor and suggested retraction. The paper was eventually retracted and published later in an open access journal with the SAME conclusions and comparisons. This was all the more upsetting since we’d been in discussions with the author(s) about reworking their data in collaboration.
I have two primary concerns and requests:
1) There appears to be a flavor in the air of “commercial chemistry software” is a bad thing. This is certainly true on many blogs discussing open source solutions. It appears to permeate into the publishing process based on some of my experiences (NOT all I should emphasize!!!!). This is my chosen career….I’ve done the PhD, a postdoc with a government laboratory, worked at a University and even in a Fortune 500 company. I eventually ended up at a chemistry software company. This is how I pay my bills, clothe my kids, live. I work with 140 other people producing chemistry software. It is a respectable career. We do science for a living. We get paid for our science by selling our products. No grants, no large company above us to carry us…our efforts are all we have to produce our living. We do EXCELLENT science and this should be the measure by which we are measured, not the fact that our science is eventually commercialized.
2) There IS good science in Open Access journals of that I have no doubt. I hope that part of the Open Access Journal process is that they check for the refusal of publication of submitted works from other journals prior to acceptance. Then they only have the ethics of the submitters to deal with. Peer-review is still necessary in Open Access Journals. My concern is with the burden of scientists who are generally overwhelmed. I get too many publications to review as it is. I have colleagues who get one a month. These are people who are already overwhelmed with work. I have no solution to the problem….just acknowledging it.
These are just observations and I welcome feedback. If and when we publish about ChemSpider where will we publish? We already have had invitations to publish in an Open Access journal…a decision is yet to be made (and a manuscript written).