4
$\begingroup$

I'm studying a TCGA dataset trying to find correlations between gene expression and clinical data which might shed light on some pathways. One column of the clinical data provides a list of chemotherapy drugs the patient received and I'm pretty sure there's a wealth of possibility in this column but I don't know anything about how chemotherapy drugs work or what kinds of things it makes sense to classify about them.

I'm making a spread sheet that will have the following columns:

Drug_Name      
Number_of_Patients_Received      
Mechanism      
Associated_Drugs      

But I don't know what else is important to know about these drugs. What would you think is important to know to help understand how these drugs might have played an interactive role with the pathways. Also, if you have any recommendations of online resources, I'll appreciate you mentioning them with a short explanation of what I might find there. But my primary question is else should I look up about these drugs to help me understand their role/impact.

$\endgroup$
3
$\begingroup$

There's this database of genetic/drug interactions, which I think is pretty much exactly what you're looking for.

Probably your best approach is to classify your drugs into a couple different categories of action. Some drugs will have to be in several categories at once(I'm lookin at you anthracyclines) Dose-dependent mutagens(cisplatin), growth-dependent mutagens(fluorouracil), topoisomerase inhibitors, tyrosine kinase inhibitors, microtubule inhibitors, retinoid signaling, immune adjuvants, and then 'misc'.

Alkylating agents and platinum-containing drugs like cisplatin directly damage DNA, while there are other drugs(methotrexate,fluorouracil) destroy or damage DNA by pretending to be nucleotides and breaking things that way.

Anthracyclines bind into DNA strands damaging it, directly create free radicals, disable topoisomerase II, and displace histones.

There are the topoisomerase inhibitors which prevent cell growth by preventing DNA from being unwound for replication.

There is a fairly diverse(chemically) class of 'mitotic inhibitors' that screw up microtubule formation or proper microtubule dissolution, any of which prevents mitosis.

Sunitinib blocks tyrosine kinases, which are involved in blood vessel formation(and nerve formation and some other things). Other tyrosine kinase inhibitors are more specific to vascular endothelial growth factor(VEGF) receptors, but most tyrosine kinases target more or less specifically VEGFR and related receptors. Some are antibodies, some are just small molecules that bind TKRs.

There are a few retinoid-related drugs most of which influence the vitamin A pathways, like tretinoin, bexarotene, etc. Dual use in skin creams.

Immunoadjuvants are designed to activate the immune system to aid in tumor cell destruction. Interferon-alpha, imidazoquinoline, et cetera.

In the miscellaneous category: Bortezomib, a short peptide that interferes with proper protein degradation. L-asparaginase slows all protein production by destroying asparagine, an amino acid. Some blood cancers respond to corticosteroids, and some ovarian/testicular cancers are treated with sex hormone inhibitors. Nearly any blood-borne poison works as a chemotherapy agent due to the EPR effect, so heavy metals or radioactive isotopes also are sometimes used. Arsenic trioxide, for instance, is just arsenic.

It's also worth looking at the steroids that are given along with cancer treatment(dexamethasone) to mitigate side effects. As regulatory hormones, there's going to be a hefty impact on genetic regulation from regular steroid injections. If nothing else it's a confounding effect you'll need to control for.

The preparations/solvents of chemotherapy agents can also be significant. Cremophor was used to suspend some anticancer drugs, but recent developments imply that might be on its way out.

$\endgroup$
  • $\begingroup$ Thank you, this looks very useful. I expect I might have more questions once I've had a chance to digest everything you've said; do you think its better to post follow-up questions here or as a new question? $\endgroup$ – Slavatron Jul 10 '14 at 17:36
  • $\begingroup$ Probably depends on the follow-up question you have. Clarification should probably go here, related but distinct questions should be new questions. I'm not a stack exchange protocol expert, so.... $\endgroup$ – Resonating Jul 10 '14 at 18:32
1
$\begingroup$

Resonating provided a good general classification scheme that could be implemented for categorizing chemotherapeutic agents. Another resource that could be used is the following: http://www.genome.jp/kegg/drug/. This is a pretty good classification of drugs based on ATC resource (http://en.wikipedia.org/wiki/Anatomical_Therapeutic_Chemical_Classification_System). These two resources are good starting points.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.