Writing
software for science
Ed
Griffiths, Wellcome Trust Sanger Institute
United Kingdom
Partner
away, so up just before 7 to sort out domestic stuff: kids
swimming, music, rugby trip tomorrow, dinner money etc...most
complicated part of the day logistically and patience wise.
Drive to the Sanger Institute through Spring countryside just
bursting into life again, the drive a small oasis of calm.
Get coffee, talk to colleague about why code won't build on
the Mac and what to do about it.
Quickly reduce incoming emails to sensible number while reminding
myself of the essential goals for today. Get on providing
screen dumping and printing facilities from current software,
trivial but so vital for those conference screen shots. While
doing this discover and fix an "off by one error",
whether you are counting fence posts or DNA sequence bases
its amazing how easy it is to make this mistake.
While doing this also think through again about our approach
to simultaneously displaying multiple sections of DNA from
different organisms in a way that hopefully will actually
be helpful to annotators. Investigate and respond
to a couple of bug reports from users of some older
software which I maintain at the Sanger.
My job is to write software to help geneticists and bioinformaticians
to annotate DNA, i.e. show where genes and other features
are positioned on the DNA, add references for supporting experimental
work and so on. This involves understanding a substantial
amount of sequence terminology and bioinformatics techniques
and principles. It involves communicating with geneticists
in many other laboratories around the world and responding
to their queries and requests about our software.
The work also increasingly involves discussions with other
programmers in the field about how to exchange information,
the semantics, the format, the programming languages, the
types of computer that will be supported and so on. Just
about everything we work on is public which is very
refreshing in comparison to work for a previous employer where
almost everything was not. Its not unusual to spend a week
at another laboratory working with someone else in the field
to "get over the hump" of a new piece of code which
will then be shared generally between researchers.
Funny to reflect how in a way I've come full circle.
I started out as a biologist (ecology in
fact), then moved in to commercial computing
for many years, now here I am somewhere between the
two!
DNA sequencing technology has appeared since I was a biologist
but my background has really helped in understanding the principles
behind it, how sequencing is done, how candidate genes are
identified and so on. This is important because the programming
here is very applied, most of the users are in the next room
or certainly just down the corridor and emails/visits to discuss
problems/improvements are frequent.
It has been a real bonus having experience of both
fields but it requires constant work to keep up to
date with the changing techniques and environment of both
programming and the world of DNA sequencing and bioinformatics.
(needs editing) |
|
