Key Dates

Related Announcements

• December 20, 2024 - Notice of Special Interest (NOSI): Molecular and Cellular Computational Tools Supporting Fundamental Neuroscience Research in Health, Mental Illness and Developmental Processes. See Notice NOT-MH-25-045.
• December 18, 2024 - NIH Research Project Grant (Parent R01 Clinical Trial Not Allowed). See NOFO PA-25-301. 
• December 18, 2024 - NIH Research Project Grant (Parent R01 Basic Experimental Studies with Humans Required). See NOFO PA-25-303.
• December 18, 2024 - NIH Research Project Grant (Parent R01 Clinical Trial Required). See NOFO PA-25-305.
• November 06, 2024 - Assay Development and Screening for Discovery of Validated Chemical Hits for Brain Disorders (R01 Clinical Trial Not Allowed). See NOFO PAR-25-063.
• November 06, 2024 - Development and Application of PET and SPECT Imaging Ligands as Biomarkers for Drug Discovery and for Pathophysiological Studies of CNS Disorders (R01 Clinical Trial Optional). See NOFO PAR-25-036.
• November 29, 2021 - Drug Discovery For Nervous System Disorders (R21 Clinical Trials Not Allowed). See NOFO PAR-22-032.
• November 29, 2021- Drug Discovery For Nervous System Disorders (R01 Clinical Trials Not Allowed). See NOFO PAR-22-031.

Issued by

National Institute of Mental Health (NIMH)

Purpose

The purpose of this Notice of Special Interest (NOSI) is to encourage the use of artificial intelligence (AI)/ machine learning (ML) methods to accelerate any of the steps of preclinical Drug Discovery (DD): target identification, lead identification, and lead optimization. 

The focus of this NOSI is on preclinical drug discovery. Investigational New Drug (IND)-enabling studies, scale-up for manufacturing, and clinical research and development are out of the scope of this NOSI.

Team science approaches where the strength and knowledge of multiple individuals across computational sciences, biology, and clinical expertise in psychiatric diseases, among others, are strongly encouraged.

For this NOSI, AI/ML refers to AI and its subsets (machine learning, deep learning, neural networks, natural language processing).

Background 

Over the last several years, the NIMH Division of Neuroscience and Basic Behavioral Science (DNBBS) program has supported the discovery and development of new drug candidates targeting different aspects of the complex biology of mental illness. Despite notable successes, such as creating a robust portfolio of new preclinical and clinical drug candidates for diverse therapeutic targets, there remains a need for de-risking and accelerating key steps of the drug discovery and preclinical drug development process. Driven by the rapid growth of big biomedical data (see, for instance, the research tools and reference data developed through PsychENCODE and the BRAIN initiative, such as BICCN and the Informatic Program), increase in computing power and continuous optimization of computing algorithms, AI/ML methods provide opportunities to expand the efficiency of discovering and developing safe and effective drugs

The preclinical DD process is iterative, multifaceted, and complex; it requires (a) basic science research and target identification, (b) target pharmacology, (c) lead identification, and (d) lead optimization and candidate selection. AI/ML programs can be applied in all preclinical DD steps. AI/ML algorithms can interpret complex biological data, predict molecular interactions, analyze genetic, genomic, and proteomic data to pinpoint potential disease targets, identify and validate suitable drug targets, predict the interaction between molecules and target proteins, help in designing drugs with enhanced specificity, potency, and minimal potential adverse effects, expedite the optimization of lead compounds and identifying potential drug candidates. Also, AI/ML methods can help predict feasible synthetic routes for the preparation of drug-like hit or lead molecules.

Research Objectives                                                                                                                                                     

This NOSI takes advantage of the rapid expansion of AI/ML methods and their application to some of the most challenging, labor-intensive, and costly aspects of psychiatric drug discovery and preclinical drug development. The central goal of this NOSI is developing and using AI/ML methods to accelerate drug design and optimization for novel psychiatric disease targets. Another goal of this NOSI is to create advanced open-source analytical tools that will be made available to researchers in academia and biotechnology and pharmaceutical companies. 

Examples of computational models may include, but are not limited to:

Target identification:

• Identification of novel targets.
• Computational models for the prediction of a target’s role in disease.
• Computational models for the prediction of a target’s role in disease.
• Analyze data from patient samples in healthy and diseased states to generate novel therapeutic targets.
• Identify preclinical pharmacodynamic biomarkers to validate the mechanism of action of the hit/lead.
• To identify genetic features associated with the hit/lead response on cell lines or organoids.
• Elucidate converging downstream pathways and potential new target identification.

Lead identification:

• AI/ML models for new drug design, including combinatorial libraries.
• Design of in silico compound libraries
• Virtual screening for novel compounds with desired drug target binding activity, hit/lead generation and optimization, drug response, and synergy prediction.
• Prediction of the Physicochemical Properties
• Prediction of the ADMET Properties
• Prediction of the druggability of targets.
• Prediction of structure-activity relationship (SAR)
• Prediction of the binding affinity and other pharmacological properties of molecules.
• Development of phenotype-based virtual screening.

Lead optimization:

• Prediction of the Dose and Toxicity.
• Prediction of DMPK.
• Modeling protein-protein interactions (PPIs) and drug-target interactions using empirical structural data or predicted structural information.
• Integration of AI/ML with 3D protein structure information for docking simulations.
• Enhancing the synthesizability of designed molecules:
  • Retrosynthesis Prediction
  • Forward Reaction Prediction
• AI/ML-guided blood-brain barrier permeability prediction and implementation into therapeutic discovery for psychiatric diseases.
• AI/ML models for structure prediction and the design of small molecule modulators of targets modulating key brain transcription/translation processes as:
  • DNA methyl transferases (DNMTs).
  • Ten-eleven translocation proteins (TETs)
  • RNA-binding proteins (RBPs) associated with translation.
• AI/ML models for structure prediction and the design of small molecule modulators of targets modulating RNAs.

Applications must include experimental testing of the predictions made by the model.

Applicants should follow the Notice of NIMH’s Considerations Regarding the Use of Animal Neurobehavioral Approaches in Basic and Pre-clinical Studies, NOT-MH-19-053.

Areas of Low Program Priority

• AI/ML computational methods not directly related to preclinical development.
• Applications proposing only the development of computational methods without experimentally testing predictions made by the model.
• Applications proposing to examine the functional effects of risk genes without high statistical confidence.

Studies focusing on computational tools and models for molecular and cellular mechanisms underlying brain processes should consider NOT-MH-25-045.

Application and Submission Information

This notice applies to due dates on or after January 25, 2025, and subsequent receipt dates through January 8, 2029. 

Submit applications for this initiative using one of the following notices of funding opportunity (NOFOs) or any reissues of these announcements through the expiration date of this notice.

• PAR-22-031 - Drug Discovery For Nervous System Disorders (R01 Clinical Trials Not Allowed). 
• PAR-22-032 - Drug Discovery For Nervous System Disorders (R21 Clinical Trials Not Allowed). 
• PAR-25-036 - Development and Application of PET and SPECT Imaging Ligands as Biomarkers for Drug Discovery and for Pathophysiological Studies of CNS Disorders (R01 Clinical Trial Optional).
• PAR-25-063 - Assay Development and Screening for Discovery of Validated Chemical Hits for Brain Disorders (R01 Clinical Trial Not Allowed).
• PA-25-301 - NIH Research Project Grant (Parent R01 Clinical Trial Not Allowed). 
• PA-25-303 - NIH Research Project Grant (Parent R01 Basic Experimental Studies with Humans Required).
• PA-25-305 - NIH Research Project Grant (Parent R01 Clinical Trial Required).

All instructions in the How to Apply - Application Guide and the notice of funding opportunity used for submission must be followed, with the following additions:

• For funding consideration, applicants must include “NOT-MH-25-050” (without quotation marks) in the Agency Routing Identifier field (box 4B) of the SF424 R&R form. Applications without this information in box 4B will not be considered for this initiative.

Applicants are strongly encouraged to contact NIMH Program staff when developing their applications to determine the alignment of the proposed work with NIMH programmatic priorities.

Applications nonresponsive to terms of this NOSI will not be considered for the NOSI initiative.

Inquiries

Please direct all inquiries to the contacts in Section VII of the listed notice of funding opportunity with the following additions/substitutions:

Scientific/Research Contact(s)

Enrique Michelotti, PhD
National Institute of Mental Health (NIMH)
Telephone: 301-443-5415
Email: michelottiel@mail.nih.gov