Week 4: Processing the Data
Aditya L -
Hey everyone, thanks for tuning in to another one of my blogs!
This week, I focused my efforts on creating a python script that can cycle through the 66 Rhodopsin proteins that I have found thus far. By using a batch script, which performs the simulations (each of the 66 Rhodopsins in the dataset with each of the 2 retinal ligands), I can run multiple trials to validate the results. Using this method, I have already performed one trial (in the figure below).
Violin plots are neat graphs that show the distribution of data. In the case of Rhodopsin data, each simulation produces a binding affinity. The more negative the binding affinity (i.e. -10 versus -8), the better the interaction. So a -10 kcal/mol affinity interaction would be a better conformation than a -8 kcal/mol affinity interaction. For each of the simulations that i conducted using the batch script that I wrote, I took note of the binding affinity and plotted it in the violin plot below. Here, I observed some interesting trends.
- Generally, assuming an ideal dataset, the peaks of the blue and orange violin plots should align at a ‘common area’ because the balance between (11-cis and dark-state) pairs and (all-trans and meta-II) pairs would be somewhat similar. However, as we can see in the violin plot below, the orange distribution is more negative. This means that the all-trans-retinal ligand performs better, on average, than 11-cis-retinal, indicating a larger bias in the data towards light-active states rather than dark-state proteins.
- While the distribution is more drawn out for Benzo[A]Pyrene, there is a considerable preference to Benzo[A]Pyrene compared to the two natural ligands.
Observations:
- There seems to be a bias in Rhodopsin data that has been collected over the years. This is expected due to experimental techniques that allow for easier collection of light-state and active proteins rather than dark-state proteins.
- Benzo[A]Pyrene, despite being a hydrophobic molecule with no hydrogen bonding, has a more favorable binding affinity to Rhodopsin. This may indicate that pollutants such as Benzo[A]Pyrene have the ability to bind to human eye receptors better than the natural ligands that exist in our eye proteins!
By next week, I will have more trials conducted, as well as some additional plots that show how Benzo[A]Pyrene binds to Rhodopsin’s various states (dark-state, light-state). I will also try to test these proteins experimentally. The long-term goal of my project is to identify how Benzo[A]Pyrene operates within Rhodopsin, and then to eventually test how Benzo[A]Pyrene docks to the broader family of 800 GPCR proteins.
Stay tuned for future blog posts, and thanks for reading!
Sincerely,
Aditya
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.