[mismith]

I am an undergraduate student majoring in biology and considering graduate school or medical school. Either way I intend to be involved in research and there is a lot of potential for development in these areas, particularly in data analysis.

My problem is that I don't know what type of research I will choose. I am looking for a language that will allow me flexibility in a lot of areas.

When looking at some O'Reilly bioinformatics books, they refer to Perl A LOT because it can easily interact with and analyze information in DNA databases (particularly Human Genome Project data). I don't see myself using this type of raw data in particular so Perl doesn't really seem like the best option.

One example of a problem I was involved in recently was in pathology. We were looking at protein production in certain types of cells. I will spare you the details of the experiment, but what we ended up with was images of cells where the protein was stained green. Some cells had a lot of green, some none, and many in the middle. There were also cells that were stained nonspecifically, meaning they were not producing the protein but stained green anyways. Our project was to "score" these cells based on the "green-ness," and eliminate the nonspecific green cells. We ended up doing this all manually, which was very time consuming and also probably statistically biased.

While this isn't the only type of data I would be analyzing, it is something I'd like to be able to do. What are the strengths and weeknessess of java, perl, python, or others for solving biological and medical problems? Can you suggest a language to start learning, and others that may also be useful in these areas?

Be ware, I am "a novice."

Thank you very much. I really appreciate your input.

-MS

[smeger]

Well, the "standard" for scientific computing is Fortran. Fortran makes me puke, though. Never has there been a less elegant language. But it's easy to get math-intensive stuff running quickly with Fortran. Check out the archives of Apple's Fortran mailing list for some examples of human genome stuff.

Once you've learned two languages, picking up new ones is pretty easy. I'd recommend that you begin by looking at one object-oriented language (Java, C++, or Objective-C) and one interpreted language (Perl, Python) and go from there.

Good luck in your endeavors!

[kman42]

For your green cell (GFP?) assay, you might want to look at NIH Image. It is scriptable and will do exactly what you are looking for. Many people use it gate cells by pixel intensity. There are several other good microscopy apps out there for both data acquisition and analysis, it just depends on exactly what you want to do.

Learning to program efficiently takes a long time and depending on what sort of research you are interested in may not be crucial. There are lots of good apps out there for doing sequence analysis, image analysis, genome/array analysis etc for use by scientists. Unless you are interested in an area of research that is so specific that no one is writing software for it or you are interested in bioinformatics, you might just want to wait and see. That being said, I know lots of grad students who wish they knew how to program because they want to do something slightly different than what's available. Most of them would get by quite well knowing a scripting language or Java. Start with something easy like Python and move on from there.

kman

[mismith]

Thanks for the great input.

Kman, I don't remember exactly we used, but we used a primary antibody we made, with a commercial secondary antibody. We stained pulmonary artery endothelial cells with mouse monoclonal antibodies (for the protein) in conjunction with Green anti–mouse IgG antibody. This was a project at a lab I worked at last summer, but I plan on checking out the NIH image for the future. What sort of projects/research are you working on?

I think the scripting language or Java is a good idea. I'll plan on checking out some books this week.

Thanks again.

[ 04-12-2002: Message edited by: mismith ]