Computational Core  

Introducing dkNET’s Computational Core (dkNET-AI.org)

Welcome to the future of research and discovery! dkNET has launched a new computational core (https://dkNET-AI.org), which will provide a set of AI/ML resources designed to assist NIDDK researchers in developing hypotheses and utilizing new AI/ML techniques. In addition, the Computational Core will also work with the AI/ML research community in developing and providing AI/ML models in support of NIDDK research areas.

Use Case

CloudMapper is a web service that allows bioinformaticians to efficiently conduct sequence alignments between reads and reference genome sequences. Users can elastically request resources from Amazon AWS to deploy a computing cluster with multiple configurable machines. With CloudMapper, alignment time is significantly reduced while ensuring low costs.

Computing Platform

dNET’s Computation Core will provide a platform to empower researchers to perform a diverse range of bioinformatics analytical tasks easily. This new AI/ML computational platform is powered by Texera, which supports collaborative data analysis to bridge the gap between computational scientists and biomedical bench researchers.

Texera (screen capture below) facilitates interdisciplinary collaboration among researchers from different backgrounds. Real-time updates on users’ status and activities create an environment of effective collaboration. Whether you are working individually or as part of a team, the new AI/ML computational platform will enhance research capabilities and drive impactful discoveries. 

Fig. 1 A screenshot of a workflow being edited by two collaborators using Texera

Texera offers several strengths that lower barriers for NIDDK researchers to utilize state-of-the-art AI/ML techniques and support multiple data modalities. These strengths include:

• Collaborative functionalities: The system supports powerful features such as shared editing, shared execution, version control, commenting, and debugging.

• Scalability: The engine of the system makes it capable of handling large amounts of data and computationally expensive tasks.

• Multi-Language support: Texera supports multiple script languages such as Python, R, and Java, enabling NIDDK researchers to leverage machine learning capabilities within their data analytics workflows. This flexibility accommodates different programming preferences and facilitates the adoption of advanced AI/ML techniques.

• Elasticity and reproducibility: Texera ensures the computing platform's elasticity, allowing it to adapt to various computational needs. This scalability feature ensures that researchers can efficiently handle large datasets and complex analyses. Additionally, Texera promotes reproducibility by providing mechanisms to reproduce and replicate analyses, ensuring reliable and consistent results.

• Dataset sharing: The platform allows community users to publish their data sets, and choose the collaborators to share datasets with. 

Please check the latest blog about using Texera to perform single-cell RNA sequencing analysis with R Language.

We have created a few use cases for you to explore and experience the full potential of the dkNET Computational Core. To get started, we invite you to fill out a Google Form indicating your interest. Let us know which specific data types you plan to work with and the use cases that you are interested in exploring. This will ensure that you make the most out of this new platform.