Forum:Do bioinformaticians often break molecular biologists' hearts by being the first author?

Question

[sentausa]

Recently I knew a molecular biologist postdoc who was upset since she was placed by her boss as a second first-author in a paper that she wrote from scratch. Her boss put a bioinformatician/biostatistician as the first first-author, with a reasoning to the postdoc that she would not have anything if they take out the data processing part.

And it turned out that she is not alone. From this postdoc's facebook, I learned that some of her friends admitted that they also experienced a very similar situation: being a molecular biologist, did all the lab work, wrote the paper, but put as a second author while the first one is a bioinformatician.

Apart from politics in the lab etc., I wonder if this is really quite common. If it is, don't you think it's quite unfair for the molecular biologist?

[Prakki Rama]

One of my friend's condition is reverse. He would not be given even authorship stating that he is just doing his job as a service. Seems like unfairness is existing in both ways.

[jli.ww]

Based on my personal experience as a computational biologist, I was not a co-first author even if I prioritized the candidates from the public dataset and other molecular biologists did the validation part, and when it comes to credits contribution, the supervisor said the computation is not important at all. Seems that many of the reversed situations happen to me.

Answer 1

[Biomonika (Noolean)]

I have heard plenty of stories from the other side. Without bioinformatician, there would be no analysis; without biologist, there would be no data. To me, it's all subject to the negotiation which should take place early in the project and be continuously re-evaluated.

Answer 2

[george.ry]

I think that almost universally, bioinformatics people see this from the other side. Generally you're stuck in a limbo of being a 'middle author', where the people that collected the samples and did the manual work in a single morning are in front of you ...

[harold.smith.tarheel]

Wow, your molecular biologist has mad skills, to finish a project in a single morning!

Seriously, it depends on how much each researcher contributed to the study (including intellectual contribution), and should be discussed in advance. If the bioinformatician runs a canned pipeline for RNA-Seq to produce a list of differentially expressed genes, and the molecular biologist identifies interesting candidates and validates them biologically (e.g., CRISPR knockouts, phenotypic analysis, and all the work that entails), then the molecular biologist goes first. But if the primary role of the molecular biologist is sample isolation and library prep, while the data analysis is the heart of the project, then the positions would be reversed.

[JC]

I agree with the "middle author" situation, but in many cases I consider it fair in many projects I have worked.

[dariober]

where the people that collected the samples and did the manual work in a single morning are in front of you

Not sure what you are referring to but in my experience sample collection and experiments go by weeks or even years. In a single day in the lab there is virtually nothing you can do.

[george.ry]

Well to make a cell suspension from tissue takes ~15 minutes (tissue dependent, obviously) and a staining with a panel of antibodies takes ~25 minutes. FACS sorting is managed by a core, as would be a downstream Fluidigm run, sample prep and sequencing. Hands on time for the wet-lab scientist is minimal. The analysis of that data takes significantly longer. Obviously the understanding relating to this experiment is years in the making, but if that morning's work were given to a technician, where I am, they would get reasonable billing in the author list.

I exaggerate somewhat, naturally, and I do wet-lab work myself, so I'm not blind to the other side. Thus, whilst I can sympathise with @sentausa, I can unfortunately point to several examples of 'mis-treatment' of bioinformaticians from just my circle of collaborators.

In either scenario it boils down to a lack of understanding about the work entailed from people on both sides of the spectrum, so there isn't really a panacea for it.

Answer 3

[John]

Your concern would be better directed at our archaic publish system, unique only to biology, where authors are ranked rather than respected as individuals.

[mikhail.shugay]

I wouldn't say that the archaic publishing system is unique to biology. In particle physics, for example, it is still very hard to understand what is the contribution of a person selected from the list of several hundred authors (Russian physicists working at CERN even call such papers "mass graves" :). It is the whole system of contribution statement and evaluation in academic papers which should be revised.

[John]

Hahah, "mass graves" is probably a good name for it - but still, things are substantially better in Physics, Mathematics, and related disciplines. The widespread adoption of open-access journals like arXiv.org negates some of the power of journals to impose rules on these authorship-rankings. Most Physics papers have their authors listed alphabetically. I've even seen some online journals order their authors randomly on each page-load so "Aaron Abrahams" doesn't always come first. In almost all of these papers, the contributions of the individuals is specified, as I'm sure it one day will be for Biology too.

If you just want to share horror stories around the camp-fire, I've seen summer students working tirelessly to develop protocols that make an experiment work, never to get authorship at all because "theres already too many on the paper".

I've seen Medical Doctors do nothing more than mail a biological sample, to then later insist (and win) the right to be first author on a PhD's 3-year project.

I've seen people work on a GWAS studies for 2+ years, only to find no result and be barred from publishing anything at all by their PI.

So if you look hard enough you'll find every manor of injustice operating in every direction - but rather than focusing on the symptoms, we should as newcomers to Biology's publishing culture try to push it in a more progressive direction. Honestly, the push, if it comes from anywhere, will come from the Bioinformaticians sick of being middle author, with that apparently meaning something.

[SES]

In almost all of these papers, the contributions of the individuals is specified, as I'm sure it one day will be for Biology too.

That already is the case. Most journals have a contributions section at the end that you must fill out in order to submit the paper. When you write the paper, you can write that section as you want and have co-authors edit their contributions as they see it. Also, before you submit you must agree to the journal's terms, including a statement that the contributions are true. The publishing system is not perfect, but I don't think that is the issue given that authors have to consciously write out the names, contributions, and agree to it.

[sentausa]

And your point is ...?

[Istvan Albert]

Their point is "it is not the player it is the game". It is not the bioinformatician or the biologist that takes the other person' spot - why is there just one spot (or actually two, first and last spots) that carries so much more weight.

On another note "breaking your heart" is both an endearing and saddening way to put it.

[sentausa]

Yes, but the problem is that we are (or at least that postdoc that I know is) still in the game, whether we like it or not. In an ideal world in the future, we might change this. But for now, authorship order is still deemed important by some people, right? Anyway, I agree that specifying contributions is one good solution.

And yeah, I saw that this girl was really brokenhearted and sad. I changed that into "upset" in my post, but I forgot to change the title :-p

Answer 4

[jotan]

I've never personally witnessed anything like this first-hand so it's probably not super common. But as other posters have pointed out, authorship disputes in general are very common. Learning how to stand up for yourself and defend your position is an important skill. Learning when not to dispute is just as important.

Having said that, it seems like the post-doc in your post was still a joint first author. This is probably upsetting on a personal level but on a professional level, being first-first or second-first will make no difference to her career.

[SES]

Very well stated. In my experience, there is always some disagreement about the order of authors and it is always a matter of perspective. It is hard to interpret this particular situation without any knowledge of the situation. I would say the best course is communicating these issues to other people involved and be firm but professional because one paper probably won't make your career.

I have a different opinion about the second point. I would say there is a big difference between first and second author. In some studies there may not be, but in many cases second author means you did a BAC prep or something similar and that is not going to land you a good academic position. Having a nice list of first author papers means you are independent, talented, probably a good leader, and you are adept at resolving these kinds of disputes.

[jotan]

she was placed by her boss as a second first-author in a paper that she wrote from scratch. Her boss put a bioinformatician/biostatistician as the first first-author,

I read that statement as meaning joint-first authors but I could be mistaken.

[sentausa]

Yes, I meant that they are joint-first authors. This postdoc was upset that her boss put her name at the second position of the joint-first authors. I agree that actually she's still a first author of the paper.

[SES]

Oh, I misunderstood the "second first-author" phrase because I've never heard that ("second-first" is kind of confusing). In the case of co-first authors, I would tell her to not be upset. It may be upsetting personally since she wrote the paper but it won't be viewed any differently if she were placed first or second since they are both first authors.

[Jean-Karim Heriche]

For people who care about order of authorship, there's no co-first, there's a first and a second. After all, the paper by A, B and C, where A and B are co-first authors, is always cited/known as A et al., and never B et al. or A, B et al.

This being said, given how twisted the whole system of authorship is and given that many studies now involve multidiciplinary cooperation, relying on order of authorship for hiring or promotion or anything else is just daft. By the way, this is also true for last authors.

[SES]

In my opinion, how a journal is cited in the text is not of major importance because many use a numerical system, in which case no author name is used (this in itself is a huge annoyance, especially when submitting to different journals but that is another issue). I'm saying co-first author is equivalent with being listed first. The ordering of the names is an issue of lab politics and I can understand personal feelings getting involved but no one in a hiring position would view that any differently if they know what co-first author means. I would argue that if you wrote the paper then you should be the only first author but we don't know all the details. In a case like this, if being first is important then you should discuss it with the PI and the other co-first author and work it out. Getting really upset and internalizing things is not a good solution in any situation.

[SES]

To be clear, I don't disagree with you. As I said above, there is a big difference to me in general, but in the case of co-first I think it's easily resolved and not something to get too upset about.

Answer 5

[DG]

Personally I've never been first author on something I didn't do the majority of writing and work on. When it comes to the biological work it all depends on the project but sometimes some follow-up biological experiments might take a lot of time, but little or no intellectual work in terms of setting up the experiments or interpreting the results. There is plenty of straightforward lab work that is time consuming and labour intensive, but could be done by students, tech's, etc.

If all I do is run data through an existing pipeline and hand the results back to let someone else do interpretation yeah, I only expect to be somewhere on the middle of the authors list, and I suspect that is what happens most of the time. On the other hand I've had projects where I've spent weeks or months tweaking and re-running pipelines and analyses. Some workflows take a significant amount of time programming and debugging but might not really be worth their own paper. In those cases it is often nice for the bioinformatician and biologist to at least share first-authorship on at least the first few papers of biological analyses done using that workflow.

Authorship is always an issue. But it is always an issue because we mostly don't deal with it up front. We know we should, but everyone is too worried about offending anyone else so we ignore it until something pisses us off. The project should be laid out well enough advance for authorship to be discussed up front. If significant changes happen later the issue may need to be revisited. But we should all be professionals...

Answer 6

[TriS]

it is interesting how this is a dichotomy made by two complementary fields in the battle for authorship. I agree with the previous comments that it should be decided prior writing it up but it also counts how much leverage the results have on the paper's message.

author's contribution is also key. now that lots of journals added this section in their papers we can know exactly what a person did. this helps laying out the fairness of the authorship order, which can then be interpreted subjectively, talking to lots of PIs at conferences/seminars etc I had the impression that, beside looking at first authorship, most also focus on the contribution (when they have the time to get to the paper are read it) and then draw their own conclusions. so, where it can be unfair that a person is put as second author, hopefully he/she'll be fairly recognized for his/her contribution in the apposite section.

a simple solution that i.e. I went through with a colleague in the lab is to look at the number of figures generated, the person who made the most becomes the first author, then second and so on. he generated one result figure more than I did, therefore he was first and I was second...this is pretty straightforward and pragmatic.

otherwise...it seems that tossing a coin is a valuable solution ;-P (cropped from web-page snapshot):

[DG]

Although in that case it looks like it was deciding about who would be first and who would be last. If both people are "senior" it doesn't matter too much, both count for about the same amount in terms of impact.

[Istvan Albert]

That is the weirdest authorship decision I have ever seen. Good to know such a thing exists.

Answer 7

[mikhail.shugay]

This is quite strange decision, as writing a paper from scratch assumes that the data analysis itself just confirmed the ideas of molecular biologist (and PI). I would consider giving first authorship to a bioinformatician only in case it is the data analysis which provided novel results serving as the backbone to the paper. Still I mostly agree with John and think that today bioinformatics and biology are so interconnected that the mere idea of a single "first author" is very wrong.

[JC]

Yes, I also consider this situation more about politics than real research. In my experience, who wrote the paper from scratch is (almost ever) the first author except when there are other interests.

Answer 8

[aleimba]

I've also seen it happen both ways, from a wet-lab and a data analysis/bioinformatics perspective. Yet more often in the case of undervalued "pet" bioinformaticians, see e.g. a "little" exaggerated from Mick Watson here: https://biomickwatson.wordpress.com/2014/07/21/youre-not-allowed-bioinformatics-anymore/.

Let's just say author contribution is a bottomless pit. Every group/lab/institute has their own policies, mostly based on politics and hierarchy rather than research. It would be served well if everybody could save time and frustration on these discussions and focus more on their work.

I wholeheartedly agree with @John and the comments beneath his post. We, as the "new" generation of life scientists, need to find a better solution for this mess (including peer review and its drawbacks).

In the current situation, however, this also infers that if you ever find yourself "unfairly" on a prominent author position, to speak up and argument your contribution doesn't warrant this and you rather be pushed back or only acknowledged. I definitely don't see this happen often ...

But there's light at the tunnel mainly based on the open science movement. See e.g. the recent software publication from Titus Brown's group and its authorship policy (http://f1000research.com/articles/4-900/v1. I recommend reading his associated blog post on his reasoning and especially the discussion that followed below the article: http://ivory.idyll.org/blog//2015-authorship-on-software-papers.html

Initiatives like the Mozilla contributor badges (https://www.mozillascience.org/contributorship-badges-a-new-project) and other badge initiatives (http://www.nature.com/news/digital-badges-aim-to-clear-up-politics-of-authorship-1.18443) might put a new spin on the system. And these are definitely not only relevant for bioinformatics/"software" projects.

[Istvan Albert]

well I agree with most except I don't know if adding everyone that commits to a github repository as an author brings any light to the end of the tunnel. That just makes the tunnel longer and darker. Now you have people that obviously should not be authors have "scientific" contributions to a field. The padded resumes and what goes with that are one of the endemic problems. Would padding everyones resume make it better for all of us? No, it just devalues real work and encourages "pretend" work.

[aleimba]

Thanks, Istvan, for your answer (and the edits). Obviously we're all arguing from our personal experiences, which is great because the subject is getting a broad perspective. But, I'm not very much concerned that this approach will have an impact on "devaluing" authorship contributions, I'm afraid that ship has already sailed (see @Johns comment above, Do bioinformaticians often break molecular biologists' hearts by being the first author?, and my examples, http://ivory.idyll.org/blog/2015-authorship-on-software-papers.html#comment-2281653681, to name just a few). As with any system it relies on the ethics of the people involved.

Nevertheless, I would love to live in a world with only "real" scientific contributions, but in the current situation I sadly more often see boldness/audacity win. So why not be open and work for a fine grained contribution taxonomy.

It probably boils down to how scientific careers are made nowadays. We can definitely agree on the problems of padded resumes. If your career depends on your oversized publication list, there will always be people who will take any shortcut necessary.

[Istvan Albert]

I actually find the feedback that you link to depressing. It shows with surprising clarity the inability of many lead scientists and PIs to look at the bigger picture and evaluate an idea on its own merits rather than by their personal, local, small and selfish needs.

They will say look, it worked for me in this particular case. Or even: look some people that did this actually managed to pad their resume. I am not not kidding that is one of the rationales listed on that page: http://ivory.idyll.org/blog/2015-authorship-on-software-papers.html#comment-2281653681 But it is obvious it would not work on a large scale. If a pull request had any value at all one could write a simple automated tool that finds typos in all github repositories and places pull requests. Now what? Do they deserve a medal?

What really irks me is that it is not just idealistically rebelling against a broken system. I would support that.

They are rebelling against a system while benefiting and exploiting the brokenness of the system. "Authorships are phony and should be done away with!". No, not gonna happen? Well I guess we all have another paper on our resume. It is a "revolution" where if the ideas fail and we go back to the things were it will actually benefit them. That is really how we can tell that it is not the right way to do it. There is no incentive to actually change anything. The ideal outcome for every "author" on papers like this is to keep things exactly as they are.

When a paper like that is accepted we are undermining and devaluing the work by every graduate student, postdoc and scientist that actually does a lot work over many years to get an authorship.

[SES]

I recently saw a project on github where the author had an automated hook to add your name to the contributor list if you starred the project. That is a great way to get visibility to your project but it does kind of cheapen the contributions of people that actually worked hard writing/testing/documenting the code, IMO. I agree with your last statement in those regards. I don't personally think that this all inclusive approach (anyone who commits is an author) is going in the right direction. That makes it harder to distinguish yourself if you made a significant investment of time into a project. I have strong feelings about this topic and could say more, but you summed it up quite well in that last statement.

[aleimba]

I hear you and I sympathize with your concern. I would'nt be happy if a publication, where I worked on for years, would get a bunch of, in my eyes, "non-significant" additional authors. Depending on the project and the reasoning behind it.

I also think, that some of the khmer 2.0 contributors didn't accept the author invitation, as Michael Crusoe pointed at here. Additionally, I'm guessing many of the contributors to khmer 2.0 contributed to the project without the expectation of getting an authorship out of it.

Let me rephrase, if my earlier text wasn't differentiated enough. I'm not saying granting authorship as Titus did is the holy grail (and I don't think there is one anyway) and should now be done for every science publication. It is most certainly more suitable for some fields than others, probably mostly depending on the degree of collaboration and project size (see physics). But, the problems of the current system have endured for quite a while, besides people knowing about its fallbacks, but still nobody had the guts to change something about it. That's why I applaud Titus for experimenting and trying something new, even if the only outcome is this much needed discussion. And mind you this discussion involves only a small part of the scientific community anyway ... What we need is a better evaluation system for scientific contribution besides the authorship/author order on a journal publication.

Moreover, I would always opt for a very open system instead of obscure back room deals. If the contribution taxonomy would be as fine-grained as e.g. working on the README, all the better. In an ideal world with no bad guys, where everybody would follow "specific" standards, we would'nt even have this discussion. But, IMO, we're already dealing with a broken system in many cases. And yes, the referenced horror-stories are extremely frustrating, yet happen on a daily basis.