Tuesday, April 26, 2005

chaos, databases, and minor victories

You would not believe the week I've had. Middle of last week, I was designing some oligos and discovered a major error in the Entrez database. As in, the people working on the Human Genome Project messed up royally, and named a protein wrong. The name they used already belongs to another protein (the rightful owner).

I contacted the guy who originally discovered this mystery gene, since I know him personally. I tried to convince him to correct the databases himself, since he's the expert, and he didn't seem like he really cared. He acted like the database people are idiots and it's their fault and not his problem, or something. More amazingly, he said the original name was wrong anyway, because the gene does not code for an enzymatic activity of the type it was named for.

At all. This thing is not an enzyme. And they never published the correction.

Anyway, long story short, I filed a complaint, and they are going to fix the database, so I feel like maybe that was my big contribution to The Progress of Science. Who knows, maybe my whole purpose in life was to find this mistake, nevermind if I ever publish another paper again! And I got a discount from a company that uses these databases to sell their products, since I ordered the wrong product and realized why on the day it was shipping out.

Ugh.

Meanwhile, I may have lost 6 months of work here on the wrong protein. I'm torn because I think I should have known better, but in a way it's not my fault if the databases everyone trusts and relies upon have mistakes in them. And, there were other issues, like my wonderful collaborators who refuse to give me their detailed protocols, and insist that it is too 'tricky' for me to set up here. And, because of issues with mailing things overseas, it has been hard to send the samples and have them do the assays that would have made it more obvious that something was wrong.

Actually I got the results of the assays last week, and it was obvious something was wrong, so now it all makes sense. Really should have done that sooner...

Meanwhile, my project has now fragmented into two parts. My amazingly supportive boyfriend says he's sure I'll sort them both out and turn it all into a cool story, but I'm feeling stressed beyond belief because I can't make this go fast enough to satisfy my fear. I'm worried it's all crap. There are no reagents to study this mystery protein that I wasn't meaning to work on. So I am ordering oligos (what got me into this mess in the first place!) and planning to clone the gene from my own cDNA preps, tag it, and go from there. I am definitely seeing an effect in my assay, so it makes sense to go ahead and pursue it. But, this is way up there on the 'most embarrassing things' that ever happened to me list. I still think the dance recital when I was about 9 years old was worse: the stereo system broke, so I was dancing to music only I could hear.

Kind of metaphorical, really....

Anyway, my undergrad is doing great, she actually seems to like research despite my warnings about how messed up the system is. And today when I went to get something signed, I ran into the Dean and he asked me how I'm doing and what are my plans, etc. Of course I said everything is terrific.

I keep trying to picture how my seminar is going to go, am I just going to present one half or the other of this story, or am I going to link it all up and explain point-blank what happened? I'm supposed to present a poster at a meeting soon, and I don't know if I will have this 'sorted out' by then. It is embarrassing to even think about it.

3 Comments:

At 4:06 PM, Blogger GrrlScientist said...

What's wrong with telling your audience point-blank about what happened? This might be useful for them to know, they might also be banging their heads against the proverbial brick wall without knowing why nothing is working right.

GrrlScientist

 
At 12:58 PM, Blogger Adam Solomon said...

Oy gevalt! That's the problem with any kind of large-scale database compiled from research. SIMBAD, the database I use for my astronomy research, has a lot of multiple entries for smaller stars, leading to a LOT of confusion for my project in particular, which involves searching the sky around smaller stars (using SIMBAD) for any possible companions. It's annoying to see two stars so close to each other, get excited, and then realize they're the exact same! lol. You did the best you could, though, and at least it won't happen to someone else in the future ;)

 
At 12:08 PM, Blogger Rob Helpy-Chalk said...

How often does something like this happen in field that depend on large databases? It seems like your woes should not be an isolated occurance. (I'm in the humanities, so I don't use many large databases, but my general experience is that they always have errors.)

Related anecdote: Watson and Crick screwed up their first two-chain model for DNA because they had taken information on the hytrogen bond from a textbook that was out of date.

 

Post a Comment

<< Home