Drug discovery is an immensely challenging problem. There are currently more than 9,000 untreated diseases with over 300 million people suffering from rare diseases, for which we are unlikely to develop treatments any time soon. The drug discovery process still costs an average of $2.6 billion per drug. Even then, 30 to 50% of top-selling drugs don't work for the patients which they are prescribed for. We build technology in the service of science, specifically using AI to tackle this huge unmet need and to transform the traditional drug discovery process. In this talk, I will discuss how integration of data is the foundation of which everything else is based, and describe our approach to using AI and human expertise to deliver validated unprecedented targets, and to enhance chemical drug design and precision medicine.

The lecture will focus on the development and application of generative models for creating novel compounds and for generating synthetic biological data with the desired properties.

Since the time of the Genomic Revolution, networks have been key to understanding biology in health and disease. Application of AI to biological networks is far from trivial. In this talk, we discuss the emerging field of network-based artificial intelligence and how it may transform disease understanding and target ID. In presenting potential challenges and solutions, we will further explore how network approaches may help identify targets and advance our understanding of the microbiome. 

Join us for a lively discussion among prominent pharma leaders, and learn:

Why, when & how to implement a public Cloud for your computing needs

Challenges and opportunities when setting and managing stakeholder expectations

Critical keys to success to realize the best outcomes

To learn more about RCH Solutions, visit our Virtual Booth

Hosted by Joe Donahue, Managing Director, Life Sciences, Accenture

Participants include: 

Andreas Matern, Head of Digital Translational Medicine, Sanofi

John Quackenbush, Professor of Computational Biology and Bioinformatics; Harvard T.H. Chan School of Public Health

Seungtaek Lee, VP, Strategic Partnerships and AI RWE Head of CoE; ConcertAI

Preston Keller, PhD, MBA, President & CCO, PercayAI

Philip Payne, PhD, Becker Professor and Chief Data Scientist, Washington University in St. Louis

Most large scale analysis of clinical trial data only leverages part of the picture, ignoring unstructured data and limiting findability across all the information collected throughout multiple disparate data sources.  This roundtable will discuss leveraging a cognitive platform to combine all data from multiple sources into one unified view using a single entry point to the data.

Evaluating, optimizing and benchmarking of next generation sequencing (NGS) methods are essential for clinical, commercial and academic NGS pipelines. Optimizations for speed and accuracy often require making trade-offs relative to other constraints. Join this roundtable to discuss benchmarking strategies, trade-offs, and the value of benchmarking genomics tools and applications. 

We are establishing an automated lab for cell profiling of drugs using multiplexed fluorescence imaging together with label-free quantitative phase imaging. This talk presents our progress and showcases how a small lab can implement a DevOps approach with modern IT solutions to carry out and sustain online prioritization of new data, coupled with continuous AI modeling building on hybrid infrastructure and microservice architecture. A key objective is on improving screening and toxicity assessment using AI-guided intelligent experimental design.

Images contain rich information about the state of cells, tissues, and organisms. We work with biomedical researchers around the world to extract quantitative information from images, particularly in high-content screening experiments involving physiologically relevant model systems. As the biological systems and phenotypes of interest become more complex, so are the computational approaches needed to properly extract the information of interest; we continue to bridge the gap between biologists’ needs and the latest in computational science (e.g., deep learning).

Beyond measuring features biologists specify, we extract more from images through profiling experiments using the Cell Painting assay, where thousands of morphological features are measured from each cell’s image. We are working to harvest similarities in these “profiles” for grouping genes, identifying the functional impact of cancer-associated alleles, discovering disease-associated phenotypes, and identifying novel therapeutics. Ultimately, we aim to make perturbations in cell morphology as computable as genomics data.

All novel algorithms and approaches from our laboratory are released as open-source software, including CellProfiler, CellProfiler Analyst, and cytominer.

The talk presents advances in multiparameter flow cytometry analysis using machine learning algorithms. Both t-distributed Stochastic Neighbor Embedding (t-SNE) and FlowSOM algorithms are very effective in the comprehensive analysis and visualization of multiparameter flow cytometry data, resulting in a deeper understanding of disease biology at the single-cell level. A cloud-based, high-performance compute environment, coupled with GPU processing, were deployed to overcome challenges with executing these CPU/RAM/GPU-intensive algorithms on large datasets.

The real world data hype caused high expectations, including RCTs might only play a minor role in future drug development. But  they are rather complementary to RCT data and cannot replace them. Artificial intelligence may change  drug development and time to market significantly, but will not replace past knowledge and experience. Real World Evidence generation can be enhanced by AI and is key for public health.

Rare disease patients suffer too often from long diagnostic delays and misidentified diseases. This creates a significant burden, not just for patients, but for healthcare systems. We present in this talk examples of instances where we have collaborated with researchers and hospital systems to develop novel approaches for rare disease patient identification using tools like genomics, machine learning, and NLP.

This presentation describes a novel data warehouse solution that captures and integrates 8 major public datasets of small molecule data.  With a pharmacology focus, the data are available for search, visualization of compound space, and machine learning.  

As an alternative to traditional drug discovery approaches, in silico hypothesis generation has potential to accelerate timely, cost-efficient identification of new uses for existing therapies. We’ll explore the current-state-of-the-art for mining extant data and knowledge resources, artificial intelligence methods, in order to generate and validate hypothesis concerning drug repositioning candidates.

How do you use data / digitization today to drive scientific discovery / product development?

What are you greatest scientific pain points / gaps that are not being met by digitization?

What kinds of outcomes do you believe digital tools could help you achieve?

Welcome to this discussion group on the growth of demand for HPC in scientific research. We are looking forward to a lively forum. We'll start by looking at three related topics:

- What events trigger demand in your organization? How has the current pandemic impacted resources?

- What could make scale and collaboration more accessible to more researchers?

- Share a recent experience of shifting workloads to manage HPC capacity.

In this session we’ll discuss how to provide researchers with performance and scale in genomics & research analytics, to drive results at a price point that’s economically viable on public & private cloud.

Storage solutions we’ve been using force bioinformaticists to make trade-offs between the capacity and low-cost of disk and the performance of flash. This results in complex tiering configurations that only deliver performance for a small slice of the data. In this session, we will review how advancements in technology enable VAST Data to revolutionize the cost of all-flash and allows bioinformatists faster analysis across larger datasets for deeper insights.