compoundcompanion_banner.jpg
MIDS Capstone Project Fall 2019

Compound Companion

“Currently, an estimated 90 percent of potential medicines entering clinical trials fail to demonstrate the necessary efficacy and safety, ultimately never reaching patients.”

Traditional approach

Traditional medicine follows a one-size-fits-all approach. Drugs and other therapies are designed to treat large groups of people with the same disease — like diabetes or cancer. They may factor in your sex, age, or weight, but overall, doctors base your treatment on what’s most likely to work for everyone with a similar illness.

One size does not fit all

Not everyone responds to a treatment in the same way. Some drugs work very well for certain people. Others don’t help at all or cause harmful side effects. Finding the exact drug that works for you can involve a lot of trial and error.

Future approach

Precision medicine is "an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person." This approach will allow doctors, researchers and pharmaceutical companies to predict more accurately which treatment and prevention strategies for a particular disease will work in which groups of people.

How we help

Predicting drug response

Associating an individual's composition with how well a compound will react is vital to making personalized medicine a reality. In order to improve our ability to predict drug response when considering molecular profiles, we must learn the genetic features that are associated with both success and failure.

Compound Companion is a data centric platform that leverages machine learning to predict an individual's chance of responding to a drug based on an individual's multi-omic profile and diagnosed condition. Our hope is that our tools can help put the right information in the hands of those responsible for delivering the right treatment to the right patient.

Review patient predictions

Once you have an individual's multi-omic profile on the platform, you can perform an analysis on the prediction results for the compounds loaded in your system. Our visualization tools will guide you in making the best decision possible for your patient.

The system allows you to then record real world evidence for the individual and use it to improve the prediction algorithms moving forward.

Recruit participants

If you are recruiting for clinical trials for a new drug with no pre-existing data, you can use our search tools to find individuals on the platform that have responded well to drugs either related to a condition, genomic target or cellular pathway.

Drug Performance

Understanding how well drugs perform across your cohort is crucial. Our dashboard will allow you to see top level performance statistics based on prediction results and real world evidence.

More Information

Last updated:

December 18, 2019