October 4, 2023
 min read

Evidential Learning and Advancements in Drug Discovery

In this blog post, we look at recent challenges for drug discovery and innovation in the pharmaceutical industry.

Evidential Learning and Advancements in Drug Discovery

Evidential Learning and Advancements in Drug Discovery 

In this blog post, we look at some recent challenges for drug discovery and innovation in the pharmaceutical industry. We also show how Themis AI’s Evidential Learning proves to be a valuable tool to meet these challenges.  

Over the past two decades, the pharmaceutical industry, valued at an impressive $1 trillion, has been facing some significant challenges. Despite committing nearly $80 billion annually to drug development, there has been a decline in this area. There has been a significant drop in returns on every dollar invested in research and development, dropping from 10 cents in 2010 to just 2 cents presently. A widely-cited academic study shows that for every billion US dollars invested in research and development, the quantity of newly approved drugs has decreased by about 50% approximately every 9 years since 1950. As the study indicates, this represents an around 80-fold reduction even when adjusted for inflation. Also, the process of developing a drug, from research to market launch, is quite lengthy and typically takes 10-12 years. The monetary implications of these challenges are profound, with the current R&D expenses for each drug reaching $2.17 billion, in contrast to the $1.19 billion in 2010.  

Amidst these challenges, the advent of AI in drug discovery represents a potentially momentous shift for the industry. In certain contexts, AI has markedly expedited drug discovery, enhancing the process by a factor of 15. For example, it has been shown that deep learning has enabled rapid identification of DDR1 kinase inhibitors, that is, substances that block the action of a protein that plays a key role in important cellular processes. Another example of groundbreaking discovery through AI is the AlphaFold deep learning system, created by Google Deep Mind, that is able to predict proteins’ structure with an unprecedented accuracy.

From a clinical standpoint, the capacities of AI in drug discovery look very promising, especially in the following areas:

Enhancing Virtual Screening: By analyzing vast datasets, AI speeds up virtual screening, helping researchers pinpoint potential drug candidates more precisely.

AI-Powered Predictive Analytics in Clinical Trials: AI facilitates the prediction of clinical trial results, refines trial structures, and pinpoints specific patient groups for enhanced success.

The AI-Driven Shift in Drug Repurposing: AI fast-tracks the repurposing of drugs by spotting existing ones that might work for new medical conditions, conserving both time and resources.

Generating New Molecules with AI: Through de novo design powered by AI, novel molecules with the desired characteristics can be created, broadening the horizons of drug discovery.

The pharmaceutical industry has witnessed a Fundamental shift in the AI landscape resulting in a wide adoption of AI tools across biopharma R&D; and companies now demand products that are tech-centric, modular, bio-specific, and secure. Benevolent UK, an end-to-end AI augmented drug discovery company, in its latest investor report points out how there is a “clear, growing market demand from Biopharma to leverage AI in drug discovery and increase the probability of success’ as a key drive for revenue generation and value creation.”  Also, Industry titans such as Pfizer, GlaxoSmithKlein and Novartis are not simply observers standing by but are proactively cultivating their AI capabilities in-house.

Themis AI stands at the forefront of this revolution and meets the demands of this industry. Anchored in more than five years of comprehensive research at MIT CSAIL, Themis AI has spearheaded innovations in uncertainty estimation, proving instrumental in molecular property prediction and discovery initiatives. In 2020, we developed evidential deep learning. This approach introduced a novel methodology for uncertainty quantification in neural network-based molecular structure-property predictions, achieved without increasing computational demands.

Molecular property prediction involves using computational models to predict specific properties or behaviors of molecules based on their chemical structure. These predictions can be about various characteristics of the molecule, such as solubility, toxicity, binding affinity to specific proteins, or other physicochemical properties.

As showcased in an academic study written by some of our team members, although neural networks excel in achieving state-of-the-art performance in numerous tasks related to molecular modeling and predicting structure-property relationships, they often face challenges when it comes to generalizing to examples outside their training data, have limited ability to efficiently learn from small amounts of data, and tend to generate predictions that lack calibration.

Hence, it is crucial to gain a deeper insight into the predictive confidence of neural models, especially in contexts like drug discovery and virtual screening, where the accuracy of model predictions plays a vital role in guiding safety-critical experimental processes. Themis AI's evidential learning delivers uncertainty estimations to allow the reliable adoption of these models in the chemical sciences. These estimates track the robustness of the models and should not be confused by the probabilities output by the model itself. 

When it comes to estimating epistemic uncertainty, evidential learning outperforms competing techniques, such as Bayesian neural networks and sampling-based approaches (e.g., model ensembling, dropout sampling). These methods only provide rough estimations of uncertainty through the use of stochastic sampling. The methods also result in increased computational expenses and longer processing times. This presents a considerable obstacle when trying to employ these epistemic uncertainty estimations for models in the chemical sciences. 

Evidential Learning offers a fast, calibrated, and scalable uncertainty estimation method that can be deployed to make models robust across a range of molecular property prediction and discovery tasks. Evidential Learning offers this solution in a model-agnostic format, without the need for sampling and or significant architectural changes.

Evidential Learning takes the concept of learning probability distribution parameters a step further by predicting distributions over the initial likelihood parameters. That is, the key to Evidential Learning is that, rather than placing priors on the network weights (a common technique in Bayesian neural networks), it introduces evidential priors over the original Gaussian likelihood function (also known as a normal distribution). In simpler terms, instead of making assumptions about the weights (parameters) of the neural network as in Bayesian approaches, Evidential Learning makes assumptions about the output of the model. These assumptions are represented as a higher-order distribution, known as an evidential distribution. The neural network in Evidential Learning is designed to learn and output the hyperparameters of this evidential distribution. Hyperparameters are parameters that govern the learning process of the model itself. In this case, they define the shape and characteristics of the evidential distribution. By doing this, Evidential Learning allows the model to express its own uncertainty about its predictions: the model is now aware of its uncertainty and thus confidence in the output.

The outcomes of this methodology are compelling. Predictions are calibrated to align uncertainty with actual errors, facilitating efficient training through uncertainty-informed active learning, and yielding improved experimental validation success rates. The numbers are quite impressive with 60% less training data required, 18% error improvement rate, and a 95% hit rate with confidence filtering. It is estimated that this could produce a 75% reduction in drug discovery costs, more than a billion dollars in savings within four major therapeutic areas and a tenfold increase in the speed of discovery.

In conclusion, as the pharmaceutical sector faces pivotal decisions and momentous changes, innovative endeavors such as AI in drug discovery and groundbreaking methodologies like evidential deep learning, hold the potential to draw the trajectory towards renewed industry growth and scientific achievements.

Evidential Learning and Advancements in Drug Discovery

Latest articles

View all

Don’t keep up with AI, stay ahead of it.

Work with us