Forum:Are we facing a bioinformatician shortage?

Question

[Christian]

There is an interesting discussion going on over at LinkedIn about the large number of unfilled bioinformatics positions and what is causing it. Anyone wants to chime in?

Disconnect between demand for bioinformatics scientists and training programs

[Devon Ryan]

The shortage seems to be fairly US-specific (unfortunately for those of us now elsewhere).

[Istvan Albert]

I think that's just a temporary situation caused by the slower adoption of sequencing instruments - they may be still too expensive for european university budgets - as the instrumentation costs drop the need will skyrocket due to various policy changes that will mandate the use of NGS in fields we haven't even heard of yet.

[Devon Ryan]

That's probably the case. The adoption in general over here seems to be slower (e.g., it seems every freaking hospital in the US is adding bioinformaticians...and that's unheard of here).

[mikhail.shugay]

By looking at some of pricing forms from various european NGS companies I was surprised that the bioinformatics analysis (e.g. quite straightforward bowtie>tophat>edger) has a cost comparable to the sequencing procedure itself. So maybe we'll soon hear about the same problem in Europe :)

Answer 1

[Dan D]

From my experience, there's definitely a shortage of competent bioinformaticians. There are plenty of people who have caught onto the demand and will advertise themselves as bioinformaticians and land positions when the gullible hiring managers don't know how to properly vet them. I've seen plenty of people with the job title who are totally lost if their favorite tool fails to answer the question. You gotta know how to code and think critically to do this job. Straight up.

If you really want to know about the shortage of competence, be part of the hiring process for a widely-advertised bioinformatics position. You'll get a lot of impressive resumes (some of them word-for-word identical sans the contact info), and just as many hilarious/awkward technical interviews (if you're lucky enough to get the go-ahead to do such a thing).

Another major problem in the field is the perception that one bioinformatician is enough to solve all the IT problems. At my last position I was in charge of server administration, desktop support (can't access facebook, help plz), application support (including deploying Galaxy on a cluster), repairing my boss' kid's feculent laptop, answering random PI data questions, giving seminars on how to do basic analysis, lab automation, website/web application development and administration, and being pimped out to other departments to curry political favor, explaining FASTQC to the 814th PI who soiled themselves because they saw a red X OMG OMG, oh yeah, and actual bioinformatics!

breathe, count to ten slowly

Note: I love bioinformatics and I think I'm lucky to be able to work in the field. But employers, if you want to attract and retain bioinformaticians, make a focused job description and stick to it once you fill the position.

[Devon Ryan]

In the land of the blind, the one-eyed man is assumed to know everything.

[Dan D]

In the land of academia, where a department of 10 employees apparently needs five administrators, there's not enough remaining money to hire a decent technical team :)

[Istvan Albert]

There is a great short story on this - H. G. Wells: The Country of the Blind - with a very different outcome

Answer 2

[Istvan Albert]

I think bioinformatics is among the most difficult of all computational type jobs. The causes are numerous and we routinely face the following:

1. incomplete specifications and requirements,
2. incomplete documentation and definitions on what tools actually do
3. frequent emergent + accidental complexities - incomplete data, incorrect data, missing key components, missing basic utility, improper experimental design
4. disconnect between expectations and deliverables, the expectations are often (always?) beyond what is possible
5. the difficulty to get the resources and recognition for a job well done, when things work out well it looks like plumbing - not particularly noteworthy

Couple this to a lower salary than that in just about any other computationally oriented field and it easy to see why the problem exists and may only get worse.

IMHO the only logical reason for a smart person to be in this field are perks that one may receive via employment: independence, no direct supervisors, prospects to moving into academia, perhaps travel - etc. And I think that is how bioinformatics works today. But it is easy to see why that may not scale up.

[smilefreak]

And don't forget the interesting problems!

[Istvan Albert]

yes good point, that too is one very important perk

[Chris Miller]

Perhaps a more optimistic slant would be to classify ourselves as biologists, but ones who specialize in computational assays. By that metric, we get higher salaries than other academics, have less trouble finding jobs, etc. :-)

[Istvan Albert]

hahaha ... oh wait .... that is plausible explanation

[Sam]

That makes everything sounds more bright and less miserable for us, haha

[mikhail.shugay]

For sure bad experimental design = one biological replica per condition:)

Answer 3

[SES]

The main issue for a lot of people is soft money in academia/research. A sane person might ask, "Why would I take a 1 year position for $50k when I can get a full time position in industry with job security and benefits AND make 2-3X more?" There are good reasons, of course. Freedom to publish, freedom to work when you want, freedom to work on what interests you. But, for people like myself that have taken on student loans from attending private schools and earning two graduate degrees, that cost is very hard to justify. You are essentially sacrificing future happiness and financial freedom for temporary independence.

I think there are plenty of talented bioinformaticians, and I have had the opportunity to work with a few that are no longer working in biology. The problem is that it is not practical to move around the world for short term contracts, especially if you have a family and/or other financial considerations.

Another issue is that biology is inherently complicated and many programmers feel frustrated by not understanding what they are doing. With so many exciting "data scientist" positions available that pay way more (and advertise fun work environments), it will be hard to get talented programmers interested in working on biological problems just because they "want to."

[Whetting]

+1 for the financial reality of academia

Answer 4

[biogirl]

Anyone see this Commentary in Nature last week?

[Devon Ryan]

That article got a lot of retweets among those of us on twitter :)

Answer 5

[Neilfws]

This is interesting from the Australian perspective where we have the opposite issue: quite a lot of bioinformatics expertise, but no available positions. Despite annual declarations regarding the importance of bioinformatics to research, no-one is willing/able to pay for it.

[Ram]

Bystander's conclusion: US industries are rich, rest of the world, including US academia, is a bit struggling for funding.

[Dan D]

I disagree that US academia is struggling for funding. I direct you to the nearest administrator/faculty party for a quick bit of perspective there. However, I do think that US academia tends to struggle with funding specific research initiatives. American universities are astoundingly balkanized and top-heavy with administrators that gobble up most of the abundant tuition, grant, and donor money.

At top universities there are a lot of grand ideas and ambitions which have too few skilled people assigned to implement the many real-world challenges. However, said initiatives do get an abundance of faculty/administrative "advisors" with all their political baggage and too little skin in the game. By the time you finally get them to agree on a specific plan they've already gobbled up most of the seed money and the initiative withers away. It's a case of too many cooks spoiling the broth. I see it again and again at so many different universities. Notable exceptions in the bioinformatics field are Penn State (The Galaxy Project is a tremendous success story), Stanford, and MIT.

Bottom line: there's enough money, but most of it is being squandered or at least not being allocated in such a way as to produce a sufficient staff.

[Ram]

I can relate to that. My current scenario is not what I portrayed. As I said, it is a bystander's perspective. But then, I suppose I have no lack of funds for justified expenses largely owing to being the go-to computer guy of my lab, much like you described in your answer :-)