This proposal is to establish an Artificial Intelligence (AI)-enabled Rapid Reverse-Translation Engine within MD Anderson. This will form the foundation of an innovative Discovery Data Science capability as part of a broader Data Science Initiative.

It will enable:

1. Discovering patient cohort-specific, multimodal disease mechanism maps that better inform drug re-purposing and novel therapeutic target identification
2. Discovering multimodal prognostic, response biomarkers or adverse-event predictors
3. Closing gaps in the translation of such discoveries from hypotheses to the clinic via innovative, AI-guided, translation decision-support

I will:

1. Develop a Discovery Knowledge-C...

Read More

This proposal is to establish an Artificial Intelligence (AI)-enabled Rapid Reverse-Translation Engine within MD Anderson. This will form the foundation of an innovative Discovery Data Science capability as part of a broader Data Science Initiative.

It will enable:

1. Discovering patient cohort-specific, multimodal disease mechanism maps that better inform drug re-purposing and novel therapeutic target identification.
2. Discovering multimodal prognostic, response biomarkers or adverse-event predictors.
3. Closing gaps in the translation of such discoveries from hypotheses to the clinic via innovative, AI-guided, translation decision-support.

I will:

1. Develop a Discovery Knowledge-Core to integrate all molecular profiling, clinical outcomes, imaging, microbiome & other data. I will initially populate it with APOLLO and selected Moonshot project data, but eventually expand it to include all such data from the MD Anderson data lake.
2. Apply my established AI technologies to automatically derive mechanistic signaling maps from the integrated multimodal molecular data and develop novel AI tools that rapidly adapt in response to changing profiles for APOLLO patients.
3. Develop and apply AI-methods to discover multimodal biomarker multimodal response or adverse event biomarkers for clinical cohorts where no biomarker currently exists.
4. For key gaps in the preclinical translation pathway, develop an AI-led decision support system to aid the effective selection of laboratory models, tools and assays.

The successful delivery of this ambitious project will deliver tangible short-medium term discoveries and form a translation superhighway to power the discovery and translation of therapies and biomarkers from valuable MD Anderson assets. This will pave the way to tailor and rapidly adapt therapy for individual cancer patients. The ever-learning knowledge-core will inform cancer therapy now, and will have the capacity in the future, to inform early diagnosis.

Read Less