Rollin' in the Office Life

Gotta say that I cannot complain about my job.

So far its been a very good week, and lots of work has been getting done.

A small review on last week:

Not much had gone according to plan, as lack of access to certain records is slowing the process to gather EB05 and EB95 values for the QT drug list. These values are basically a newer kind of statistics called Bayesian statistics, where instead of calculating the "Frequentest approach" ("if I did this EXACT experiment 100 times, 95% of the results should fall in this confidence interval") we calculate true values ("the true value in nature has a 95% chance of being in this interval"). These stats are calculations that are so ridiculous that they have only become popular due to the rise of computers. Using Bayesian Stats allows us to confidently know a ballpark estimate of REAL answers, as opposed to experimental answers that Frequentest Stats calculate. The EB05 and EB95 are simply the bounds in which the true value for something must reside.

WHAT IS IT THAT WE ARE TRYING TO FIND WITH THESE EB VALUES?!?

Well we are trying to see if this Drug A causes Torsade de Pointes. If people are getting TdP and taking this drug, then there is a possibility that it could cause this condition, but there is the risk that by coincidence that they develop the condition while taking this drug. This is where the stats come in.

If drug A DOES have some risk to cause the condition, then the risk will be some percentage value that is truly unknown. If there is no risk, then the value should be zero in nature. The EB05 and EB 95 give us a range of what the possible percent risk could be. If it encompasses zero, then it is unclear if the drug has a risk. If the encompassed values are low, then we can passably allow the drug to be issued with a warning. If the values for risk are high, then we may even pull the drug off the market. Stats are all over pharmacology.

As for what I am doing now, I am finding if certain drugs are chiral or not. This means the molecular structure has a mirror image of itself in someway. Think about your right and left hands. They are basically the same, but no matter how much you try, you cannot make one the same as the other. It just doesn't match.

Apparently chiral molecules act WAY DIFFERENT than their twins, some have opposite effects even. Some cure different diseases. This is due to the shape of the active sites for the molecule to "do work in". Its like trying to put a right hand in a left glove. My job is to figure out for all the drugs on the TdP list:

1. Is it chiral?
2. Which chiral molecule is used?

Sometimes both will be used in equal proportion, which is called a racemic mixture. These are becoming more rare nowadays, as different isomers have drastically different effects, so single chiral isomers are replacing the old racemates. Racemates used to dominate the market because they were cheaper to manufacture, but that is not the case anymore.

Food Update: I bought some orange juice for the whole office, to give back a little, and apparently no one likes OJ. So I got it all to myself. They like brownies though.