A tour of HumanBase, an interactive AI platform for human biology

Join dkNET Webinar on Friday, April 24, 2026, 11 am - 12 pm PT

Presenter: Rachel Sealfon, PhD, Research Scientist, Flatiron Institute 

Abstract

The wealth of existing heterogeneous biological datasets presents computational challenges for analysis as well as opportunities for learning valuable functional insights. Cutting-edge algorithms can facilitate analyses but are often not accessible to biomedical researchers or anyone lacking expensive computational resources. Here, we present HumanBase (humanbase.io), a data integration and analysis platform that applies AI/machine learning algorithms to learn tissue, cell type, and context-specific biological associations from massive genomic datasets. HumanBase integrates ~62,000 experiments, inferring the chromatin, expression, and post-transcriptional impact of genetic variants, predicting tissue-specific functional networks, and clustering genes into functional modules. HumanBase can guide experimental design and discovery by democratizing advanced, computation-intensive methods through intuitive web interfaces and distributed computing resources. This powerful, broadly accessible analytic platform allows users without coding knowledge to run high-performance end-to-end predictive pipelines and generate publication-quality visualizations, while the API provides programmatic access.

The top 3 key questions that HumanBase can answer:
1.What additional genes are functionally related to my gene(s) of interest in a specific tissue context?
2. What are the main function and process themes in my large list of genes?
3. What is the impact of a non-coding variant on regulatory activity, epigenetics, post-transcriptional regulation, and cell-type specific gene expression?

Dial-in Information: https://uchealth.zoom.us/webinar/register/WN_P-WOMsuARj-JIpznHJBjHQ

Date/Time: Friday, April 24, 2026, 11 am - 12 pm PT