Machine learning is a core technology that is rapidly advancing within type 1 diabetes (T1D) research. Our Human Islet Research Network (HIRN) grant is studying early cellular response initiating β cell stress in T1D through the generation of heterogenous low- and high-throughput molecular measurements, which are well suited for use as AI/ML ready datasets, as well as the development of generalized software solutions to transform disparate data into common formats with appropriate metadata for AI/ML. The data packages developed will focus formats that can address challenges in pre-processing (e.g., imputation), feature extraction, and other gaps in the processing of single- and multi-omics datasets that can be addressed through AI/ML methods for both small and large sample sizes.
We provide AI/ML ready omics datasets that are at various stages of data processing that they are amenable to (1) new methods to improve the pre-processing of single- and multi-omics datasets and (2) predictive modeling of endpoints of T1D.
Descriptions and dataset links for each of the following research projects can be found at the following urls:
- Diabetes Autoimmunity Study in the Young (DAISY): https://data.pnnl.gov/group/nodes/project/33479
- Human Isolate Research Network (HIRN): https://data.pnnl.gov/group/nodes/project/33478
- The Environmental Determinants of Diabetes in the Young (TEDDY): https://data.pnnl.gov/group/nodes/project/33477