It was started in 2001, and the first issue came out January 1, 2002. It’s published by the American Society for Biochemistry and Molecular Biology (ASBMB), which also publishes the Journal of Biological Chemistry (JBC), a rather substantial pillar of the biosciences publishing community. The JBC is the largest single biochemistry journal in the world. It was started over 100 years ago. The JBC has done a lot of important things in publishing. It was the journal that began the electronic revolution. It spawned the HighWire Press, which publishes in excess of 950 journals. The JBC was its first journal and its electronic debut was a joint financial and intellectual effort of ASBMB and HighWire Press.

I was Treasurer of the Society at the time that was going on and I was intimately involved in all the discussions. I was also an Associate Editor of the JBC. I know how they (ASBMB and Stanford University and what eventually became HighWire Press) worked closely together to make electronic publishing a reality in the biosciences. While there were some previous electronic publications, mostly in physics, I believe, this was the first really established international journal in the biosciences that went electronic. Needless to say, many publishers followed suit in fairly rapid succession. Almost all journals still have print versions but, slowly but surely, electronic versions are taking over.

All of this is important background for Molecular & Cellular Proteomics (MCP). After the substantial successes in leading the pack of getting things into the electronic world, the Society had a retreat in 2000. This was one of those affairs with a lot of introspection, and one of the subjects we discussed was our successes of not only making an electronic version of the JBC, but also making the whole submission process electronic—there’s no paper passing around in the submission or review process—and we thought, "Now that we’ve got all this technology, why not put it to good use?" There emerged a clear notion that the Society should expand its publication activities.

“We’re doing many things on the cutting edge of proteomics.”

At the time, we (ASBMB) had just taken over the Journal of Lipid Research, which had been around for nearly 50 years, published by a small non-profit organization, but we thought we were also in a good position to start a new journal of our own. The area we agreed on was proteomics. We wanted to publish in the larger spectrum of biochemical sciences, and we felt that this was an up-and-coming field and one that, at that point in time, was quite underserved, in terms of journals that were really interested primarily in it. While people were publishing proteomic material in other journals, including the JBC, there were essentially no journals devoted solely to it in 2000.

So we made the decision to start the journal. Richard Hanson, who was President of the Society at the time, asked me to get MCP started and to eventually be Editor, largely because proteomics is part of the larger field of protein chemistry, which I’ve been doing for some 40 years. The Society moved very quickly after deciding to do it, and having been an Associate Editor of JBC for 13 years, I was familiar with the entire scheme of electronic publishing used by the ASBMB, which made me a logical choice.

The answer probably depends on who you talk to. If you talk to the past treasurer, he’d probably throw up his hands, and say it was not inexpensive at all. It was certainly cheaper than it would have been if the JBC hadn’t provided the electronic background. There was efficiency in the sense of economy of scale; things were already in place that didn’t have to be done again. But, that said, the Society has not reached a break-even point on the journal yet.

The problem is that building a subscription base of an absolutely de novo journal is not easy to do. And unlike other journals of this type, it’s not bundled with anything else. In other words, ASBMB could say, if you take JBC, you take this too. They have chosen not to do this. So it’s been an uphill struggle to get subscriptions, particularly from libraries (ASBMB members get it free as a membership perk). Many other journals, including our principal competitors, are bundled. So even though we have a higher impact factor and we’ve been the number one journal in our field since we were rated, we still have a significantly lower subscription rate.

I think the quality of what’s in MCP drives citations. At least, that’s the only explanation I can offer. We adopted extremely high standards since day one. That’s a self-serving statement, but I can tell you that it’s true. With my Co-editor, Al Burlingame, we have from the get-go retained people as Associate Editors, who we felt have high standards. As a result, we’ve been very stringent in making sure, to the best of our abilities, that we only publish good material, and I think it shows.

In the early days, there was always a struggle to get enough manuscripts together to get an issue out. Back then, Barbara Gordon, who was Director of Publications (and is now the Executive Officer of ASBMB), would often plead with us because she needed more papers to put together an issue and publish close to the cover date. And certainly getting people to submit high-quality papers in the early days of a journal was a challenge. But we did. In fact, Al and I spent a good deal of our time going to proteomics meetings and, if not badgering people to submit good papers, then at least constantly trying to convince them that MCP would be a good place for their work.

I believe that the philosophy of this journal from the start has had a very significant role in the number of citations. In the proposal for forming MCP that we wrote for the ASBMB leadership, we said we did not want it to be a repository; we wanted it to be an active force in the development of this field. We feel we’ve done that.

We’re doing many things on the cutting edge of proteomics. We took the lead in developing guidelines for publishing protein identification data, particularly by tandem mass spectrometry. That’s basically two mass spectrometers in a row—usually abbreviated MS/MS. This technology drives a very high percentage of proteomics work. Easily two-thirds of all the papers we receive at the journal, in some way or other feature, or at least use, this approach. Because of that, we recruited as Associate Editors some of the most prominent people working with this technology—particularly Ruedi Aebersold and Steve Carr. These two individuals, plus Al Burlingame and myself, were the driving force to come to grips with this central problem plaguing proteomics, namely errors occurring in protein identification experiments.

Among other things, what people are trying to do with proteomics, which they didn’t do with genomics or traditional protein chemistry, is dissect and identify proteins in complex samples that have hundreds or even thousands of proteins in them—and they use this MS/MS technology to do so. The possibilities are fantastic and it’s one of the reasons why everyone is so excited about proteomics. People literally chop up whole livers, and then try to identify as many proteins within this sample as they can. Nobody has been able to do this before on this scale, but the problem is that a lot of the identifications are just wrong. Thus the literature is filling up with mis-identifications. Part of the problem is that people have been over-extending the technology. One person identifies 400 proteins; the next one wants to do 450. People have been pushing the boundaries, and they’ve pushed it to thinner and thinner degrees of reliability of the data. The problem has reached crisis proportions. Significant papers were being published that were full of errors. We decided that journals, as the gate-keepers of this information, had to do something about the problem.

We started with in-house guidelines, which were the first attempt to tell authors that their papers had to contain certain kinds of information so we can assess the reliability of what’s being reported. When we released them (which we called the Carr Guidelines because the formulating committee was chaired by Steve Carr), they created a great deal of interest. Then lots of other people started weighing in, and we realized we needed to expand them. The next step was, under the auspices of the ASBMB, to hold a two-day meeting in Paris, to which we invited about 60 people with a variety of backgrounds, and maybe 30 showed up. Considering that most were traveling on their own support, this was a pretty good turn out.

We had two days of discussion on the existing (Carr) guidelines and we expanded and revised them, producing a new draft, which we then circulated for three months to everyone we could think of, including other proteomic journals, asking for feedback. Then this same group of 30 people took the comments that had been received, revised the guidelines, assembled them into a final set of rules, and presented them at the Association of Biomedical Research Facilities meeting in Long Beach this past February. As far as MCP is concerned, these now must be fulfilled by any submissions to our journal. If you send a paper to us with this kind of data, you must comply with these guidelines.

We have not tried to assess to what extent these have been adopted by other journals. We sent them to the other proteomics journals (including ones that only publish such papers occasionally) with the suggestion that they might wish to consider them as possible guidelines, wholly or in part, for their journals. We believe that they address important issues and we feel it’s very important to tighten up the proteomics literature.

This is an illustration of the kind of things this journal has done so that we would not be just a repository of information and why we see ourselves as an active force in helping this field to develop.

It takes a year or two before you even have an impact factor and can judge how you’re doing. Our first impact factor was 8.3, and we were enthusiastic about that. The next year it went up to 9.6, which was a nice rise. Last year, it went up to 9.9. We just missed 10—I would have loved it to have been 10 but we didn’t quite make it. Nonetheless, we feel these numbers are a good achievement for a young journal.

Moreover, most journals have a high component of review articles which often accumulate a great deal of citations. We publish relatively few review articles. Almost all of our papers are research or technology articles. The latter are not methods papers but have to describe some significant advance in an aspect of technology that supports and is key to some area of proteomics. We’re not looking for someone who changed the buffer of a two-D gel so that the response is somewhat clearer. Our technology papers have to be the kind of improvements that are considered to be major steps in the development of the methodology. There are always things that you feel that you can do better but I would say that, overall, MCP has performed as well or better than we had hoped when we set it up 5 years ago.

Most people would ask me immediately what the impact factor was. I would say, "We don’t have one yet; we’re too young," and I would then say, with a smile on my face, "I’m sure it will be good when we have one." And that turned out to be the case. So what else would I say to them? I’d say, "This is an up-and-coming journal that really wants to support the field of proteomics. Since your career is now strongly tied to the development and growth of proteomics, you should support this. And you should support it because this will become a principal journal in your field." Enough people believed that that we got good papers and that in turn made MCP a success.

At the same time, a lot also has to do with the credibility of the people involved. Our Associate Editors and Editorial Board were and are well known in the field of protein chemistry and proteomics. They were all hand-picked, and have included an international spectrum of people from both industry and academia. I also think that the prestige of the ASBMB as a quality publisher was a major plus for us.

It was always less than 50%. Right now, it’s around 30 to 40%. It hasn’t changed enormously, believe it or not. I guess that’s because our standards haven’t really changed. You can’t keep a good impact factor if you publish poor papers. I did an interesting analysis you might be curious about. About six months to a year ago, I was wondering whether our high impact factor was due to a handful of papers being cited hundreds of times and the rest not cited at all, or whether the citation pattern was more evenly distributed. My view was that if it was the former, we were walking on a tight rope. If some year we didn’t publish such high impact papers, then the impact factor would likely tumble. I was surprised to find that in the first three years of publication, we had only one or two papers that weren’t cited at all. We had papers cited 70, 80, 90 times, but virtually every paper published had at least one citation. So the answer was we were being cited fairly uniformly across the board.

One of the issues that we have wrestled with, and that we’ve yet to come up with an answer for, is how to deal with raw data. We are participating in a beta experiment at the Broad Institute at MIT, where some raw data has been posted to see whether or not people want to go into the files and interrogate them to find new things. Typically, when people collect large data sets, they analyze a portion of it, write the paper, and publish it. Much of the raw data is never used. By posting the raw data so that other people can interrogate it, may allow some one else to get something out of it.

Since the beginning, proteomics has been characterized by large data sets and big experiments. The idea is how do we capture more of this data for the same amount of money? It’s quite clear that the raw data cannot be part of the journal. Even with electronic versions, it’s not feasible—these data sets run to gigabyte size. We have to figure out how this can be set up, correlated with our journal or others and still accessible to the public. I predict we’ll solve that problem. When we do that, I think we’ll see some interesting new analyses based on some already analyzed data.

We’re also continuing to assess our guidelines. We’ve now become interested in setting up guidelines for clinical proteomics papers. We have always been interested in publishing articles in this area and we have had a clinical proteomics advisory committee, composed of practicing physicians still active in doing research, since early on. As with MS/MS data, there are major problems associated with papers in clinical proteomics. For example, somebody isolates a tumor, chops it up, measures the proteins in it, and wants to publish it. But it’s an n=1. It’s not reproducible, by definition. These and many other problems are real issues in clinical proteomics.

So Julio Celis, one of our Associate Editors, who is head of the Danish Cancer Institute in Copenhagen, is setting up an international committee, and we’ll do the same thing we did with the MS/MS issue—hold a workshop, draft guidelines, and then take the lead in promulgating them. The problems are different in many cases to the ones dealt with by the protein identification group, but they’re problems we think we can handle. This is the kind of thing that will help the field develop further. Obviously clinical proteomics will remain a big part of the proteomics lure. The applications are very attractive. There is a need to get better biomarkers, better diagnostics, and eventually better treatments and proteomics will certainly play a role in these developments. MCP clearly plans to continue to play a significant role in reporting these types of findings.

Molecular & Cellular Proteomics
Ralph A. Bradshaw, Co-editor
American Society for Biochemistry and Molecular Biology, publishers