The Toxicity Atlas - One year later
It’s been about a year since I started my PhD project named The Toxicity Atlas. In this four-year project, I’m collaborating with researchers from Amsterdam UMC and my colleagues at Medstone to find better treatment options for patients with brain cancer. Let me tell you more about my results so far.
Information databases
Developing new drugs costs a lot of money and a lot of time – the journey of developing a drug from scratch can take up to ten years until it can be sold on the market! That is why, during my PhD, I am researching the following hypothesis: can we give patients combinations of existing drugs that will work synergistically, to create better treatment options against cancer? In order to answer this question, I must first find out what information there is available about these existing drugs. So, I’ve spent the past year digging deep into drug databases and I’ve noticed there are three types of drug information:
-
The chemical properties of a drug
-
The effect a drug has on cell cultures and animals (preclinical research)
-
The effect a drug has on humans (clinical research)
Each type of information can be found on chemical, preclinical and clinical databases, respectively. I need all three types of information to be able to predict which drugs will work well together and create a better treatment.
Searching for data
It turns out that there aren’t many drugs that appear in all three types of databases. I did find a lot of data about the effect of drugs in clinical trials (e.g. the effect a drug has on humans), but surprisingly enough, not many of those drugs show up in the preclinical database (where I was hoping to find info on how cell cultures and animals react to the drug). This didn’t make sense, because (as the name suggests) drugs have to go through the preclinical phase before they can be researched clinically. There appears to be a lack of information about preclinical research, at least in publicly available databases.
So I went looking – where could this preclinical data be? It turns out that these preclinical results can’t be found in the databases but in scientific journal publications. This is good news because this information can be useful in my drug predictions! The downside is that quite some work needs to be done before I can use this data. To predict whether two drugs may form a feasible combination, I’m using a computer model. This model needs the input data to be in a certain format – like the data we find in the databases. That means I will have to extract the data from the scientific articles and convert them to another format before the computer models can use them.
With help from SURUS
If I would go to extract and analyze the information from thousands and thousands of preclinical publications myself, it would take years! Luckily, I can use SURUS to take a lot of work off my hands. SURUS is trained to extract information from scientific publications, so that’s exactly what it’ll be doing! This way, I am trying to fill the information gap in the databases with data from scientific publications.
Once I’ve gathered all of the information, I can continue my research and look for possible drug combinations to treat brain cancer. If you’d like to stay up to date with my PhD project, you can follow Medstone via LinkedIn.
The results of this research are yet to be published. See also: Scoarta et al. (2022) A roadmap to use non-structured data to discover multi-target cancer-therapies. Article in preparation.
-
Most popular related searches
Customer comments
No comments were found for The Toxicity Atlas - One year later. Be the first to comment!