[Baypiggies] Bioinformatics Python Programmer

[Andrew Dalke]

Hi all,

I know nothing about this particular job, but a few things about what  
it's like to do software development for a bioinformatics/ 
cheminfomatics/structural biology group, so I thought I would add  
some feedback given the other comments I've seen.

On Oct 6, 2009, at 3:36 AM, Glen Jarvis wrote:
> So, here are some basic questions to help you sort out, "should I  
> even read on?"

What he's asked starts with high school biology and builds up to  
undergrad microbiology/bioinformatics. What he's asking for is  
someone who knows the basics of the bioinformatics. (If you noticed  
that his questions show more a bioinformatics, vs. structural biology  
slant, then you're definitely qualified. ;)

I've worked with people who have no biology/chemistry background when  
entering this field as software developers. It can happen. But it's  
rare. In general it's easier for a scientist to learn enough  
programming than for a programmer to learn enough science. Without  
the domain knowledge it's hard to even ask the right questions.

There are cases where specific abilities, like optimization, outweigh  
knowing knowledge of the field, but this posting is obviously not  
about one of them.


> What would work in our team, you may ask?  Someone incredibly  
> proactive and who won't be intimidated easily. For example, if a  
> PhD told you they needed something to happen, described it  
> graphically (what it should look like), but not understand any of  
> the difficulties (like databases, javascript, html, etc.) to make  
> that happen, would you get frustrated, or would you make a plan on  
> your own and make that happen? How could you deal with the  
> frustration if the PhD didn't understand what was taking so long  
> because they're focused on the result instead of infrastructure,  
> how would you address this proactively?

Those situations are very common, and he describes it well. It might  
be common as well outside the sciences, but I don't have enough  
experience in the wider world.

I had one of my clients say to me (humorously) when I was describing  
the different technology solutions "you're saying things but all I  
hear is a bunch of sounds."

This is part of the work I find fun, btw. Your customer doesn't want  
to be involved in making the technology decisions, just figure out  
what they want and get the work done.

> Can you read incredibly messy old HTML, JavaScript and Perl code,  
> but write incredibly pristine python/django/jquery code, and do it  
> quickly?

For some background on this in general (again, might have nothing to  
do with this particular job) see the part where I said "easier to  
teach a scientist to program"? Well, most of them don't program well.  
They learn just enough to get the science done. They've spent ~6  
years of graduate school, plus industry experience, to get where they  
can do science. Not code.

> Are you willing to take an incredibly low salary (comparatively)  
> because you're that interested in science, working for an very well  
> known research facility, etc?

That's the tricky part for many people. When I got started in this  
field as a software developer I was cautioned by a recruiter that the  
wages were lower than in the general programming field. For one, why  
should someone with an undergrad degree at best be paid a lot more  
than someone with ~6 years of graduate school, plus industry experience?

(There's also something which has only happened to me a couple of  
times in my career - looked down upon because I didn't have a PhD.  
There's a cultural tradition I've heard of in some places where the  
PhDs rule and everyone else is a bunch of lab monkeys. I've not come  
across it; it might be more common in a wet lab environment. Then  
again, most people assume I have a PhD.)

But there are people, like me, who like working with more physically  
oriented data sets than working on, say, financial data or social  
networks.  People who just like science, and will take the pay cut.

It's also great fun to work with people who have absolutely no  
problem in peering intently at data and spending hours trying to  
figure something out. They want to do the things you're helping them  
do, and you usually get quick feedback.


BTW, I'm available for consulting and contract programming in the  
life sciences, and programming training so those scientists can be  
more effective at doing their science ... but I charge pharma rates  
and don't plan to be an employee. ;)


				Andrew
				dalke at dalkescientific.com