Monday, February 15, 2010

"Is there a palaeobiologist on board?"

"Is there a doctor on board?"

This is one of those lines you hear in movies, dramas, cartoons, etc. but you never expect to hear in real life. But recently I actually heard this announcement on an airplane, though unfortunately I don’t know what the situation was, nor did I have the chance to find out if there really was a doctor on board, if he/she came forwards, and what happened afterwards. So no story there…

But I did have a thought:

“What if they wanted a PhD instead of an MD? And what kind of emergency situation would require the services of a PhD in palaeobiology?”

Perhaps the following:

Announcement: Is there a doctor in palaeobiology on board?

Palaeobiologist (PB): Why yes, I happen to have a PhD in palaeobiology.

Flight attendant: Thank goodness. Please follow me to the cockpit.

PB: Certainly.


PB: What seems to be the problem?

Pilot: We have a terrorist situation here and we need your expertise.

PB: Umm…I don’t think I’m qualified to deal with terrorists…

Coilot: Oh, you’re more than qualified.

Pilot: You see, there are explosives on this plane and the only way we can prevent them from being detonated is for us to comply to a series of commands. The terrorists sent us a data matrix and a phylogeny, and we have to analyze the data following their protocol and get the correct results within the remaining flight time. That would defuse the bomb.

Copilot: But we keep getting an error. We simply followed what they did but it just doesn’t seem to work.

Pilot: I’m afraid this is beyond us.

PB: I see…may I?

Pilot: Of course.

PB: Ah, the reason for the error is because you’re using the whole dataset; you have more rows in your matrix than there are tips on your phylogeny.

Pilot: …meaning?

PB: meaning, there is a discrepancy between the numbers of entries in the data and the phylogeny.

Copilot: Right, but a lot of the data entries seem to be duplicates.

PB: Actually, they’re not duplicates because the data is on the specimen level. So each species can be represented by more than one row.

Pilot: So how do we deal with this?

PB: Not to worry, I know a sorting code that will sort out the data so that we have one datum per taxon. According to their protocol they took mean values, so we just compute the mean for each species. But you’d also have to realign the data so that it is in the same order as the tips of the tree. I have a code for that too.

Pilot: Why didn’t they write all this in their protocol to begin with?

PB: Well, it never really occurs to many people that what they think is common sense may not necessarily be for others.

Copilot: Damn, terrorists!

PB: There you go. Now if you rerun the analysis, it should give you the correct results. But since the tree is huge, it may take a couple of hours to run.

Pilot: Thanks very much; I don’t know what we would have done without you.

PB: No worries, just glad to be of service. By the way, how did you know to ask for a palaeobiologist instead of a computer scientist or a statistician?

Copilot: I noticed that the list comprised of Linnean binomials, and I noticed further that they represent fossil species

PB: I'm impressed; how do you come to know fossil taxonomy?

Copilot: I took a basic palaeo course as an undergrad.

PB: Good man!

……but of course the crew could have asked for an evolutionary biologist instead because those guys tend to be more numerical than us!