# Neural computations underlying the relationship between social learning and mood

> **NIH NIH F30** · ICAHN SCHOOL OF MEDICINE AT MOUNT SINAI · 2024 · $50,374

## Abstract

PROJECT SUMMARY/ABSTRACT
Social interaction strongly influences both momentary mood and general mental health. Indeed, social isolation
during the recent pandemic caused record-high levels of mood symptoms in many individuals around the world.
This relationship between social interaction and mood has been documented not only in mood disorders, but
also in the descriptions of many other psychiatric illnesses. So, better elucidating the influence of social
interaction on mood may have powerful implications for treating those burdened by socio-affective deficits.
Recent work in computational psychiatry has defined mood as a progressive impact of reward prediction errors
(RPEs), which reflect the differences between predicted and perceived rewards. These efforts have been
expanded to account for learning specifically within the social domain. In society, we must often learn social
norms, or agreed-upon rules of behavior among social groups. These norms change over time, as driven by
norm prediction errors (nPEs), which act as learning signals that help us to predict norms more accurately in the
future. nPEs, like RPEs, influence short-term mood. Yet, it is not well known how nPEs relate to long-term mood
symptoms. We are also limited in our understanding of how nPEs are encoded in the brain. Prior work has
predominately relied on neuroimaging techniques, which can identify high-level cortical circuits involved in norm
prediction but cannot as reliably assay the potential involvement of subcortical micro-structures. Alternatively,
human single-unit recording may be used to directly examine the roles of micro-structures in norm prediction at
the resolution of single neurons. Single-unit studies have identified neurons in the substantia nigra (SN), a
subcortical nucleus that is hard to reach with fMRI, that are involved in RPE encoding. Given similar
computational relationships with mood and associated activity on neuroimaging, RPEs and nPEs may also share
common encoding by SN neurons. To date, this hypothesis has never been tested. Considering these critical
knowledge gaps, the objective of this proposal is to investigate the neurocomputational mechanisms underlying
the influence of social interaction on mood by leveraging computational modeling (Aim 1) and human single-unit
recording (Aim 2). The central hypothesis is that nPEs influence both short- and long-term changes in mood; we
further predict that nPEs are neurally encoded by single-units in the human SN, which is supported by pilot data
collected from twelve patients so far. These findings will expand our neurophysiological understanding of the
relationship between mood and social learning in health and disease in hopes of informing future
psychotherapeutic interventions for disorders that adversely affect social and affective functioning. This research
will take place at the Centers for Computational Psychiatry and Advanced Circuit Therapeutics at the Icahn
School of Medicine at Mount Sinai. ...

## Key facts

- **NIH application ID:** 10996525
- **Project number:** 1F30MH135658-01A1
- **Recipient organization:** ICAHN SCHOOL OF MEDICINE AT MOUNT SINAI
- **Principal Investigator:** Arianna Neal Davis
- **Activity code:** F30 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $50,374
- **Award type:** 1
- **Project period:** 2024-07-26 → 2026-07-25

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10996525

## Citation

> US National Institutes of Health, RePORTER application 10996525, Neural computations underlying the relationship between social learning and mood (1F30MH135658-01A1). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10996525. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*