# Predicting and controlling polygenic health traits using probabilistic models and evolution-inspired gene editing

> **NIH NIH DP5** · UNIVERSITY OF CALIFORNIA BERKELEY · 2024 · $401,250

## Abstract

Predicting and controlling polygenic health traits using probabilistic models and evolution-inspired gene
editing
PROJECT SUMMARY:
New mutations are a source of adaptive evolutionary novelty but can also cause genetic diseases and cancer.
While we can now correct detrimental mutations using CRISPR/Cas9 technologies, DNA modifications can have
unintended consequences through seemingly unpredictable epistatic and environmental interactions, as could
well be the case for the presumed HIV-resistance mutations in CCR5 recently CRISPRed into humans. In higher
eukaryotes, fitness or health traits such as adaptability or disease susceptibility appear to be controlled by
numerous mutations acting in concert – they are so-called polygenic or complex traits. Such mutations might
even manifest detrimental in some environments while beneficial in others, therefore also called antagonistic
pleiotropic. The main goal of the proposed work is to use the versatile model plant Arabidopsis thaliana to
enhance the predictability and control of the polygenic and antagonistic fitness effects of mutations.
Results from this project will provide universal principles to deepen our understanding of complex human genetic
disease and inform the safe correction or avoidance of harmful mutations in the future.
 Specifically, I will pursue the following aims: 1) predicting polygenic fitness effects across
environments, 2) improving fitness by controlling deleterious and beneficial mutations using
multiplexed genome editing and mutator alleles. Arabidopsis thaliana is an ideal model to tease apart the
fitness effects of mutations in complex environments due to its high malleability to engineered mutations, and its
extensive community and resources. The 1001 Arabidopsis Genome Project and a genome-wide Knock-Out
(KO) collection allow for quantifying fitness of thousands of publicly available natural and artificial mutations
across environments. Building a global network of Arabidopsis researchers, we have started an experiment with
the same natural strains in 45 locations, which I will use to quantify environment-associated mutation effects.
Integrating this with information of relevant KO lines, I will build on my previous predictive models to understand
the effects of mutations on fitness across environments, and the features that make them deleterious. Such a
deep understanding of mutation effects will ultimately allow us to alter fitness in predictable ways. I will test this
in two ways: First, using multiplexed CRISPR base-edits, I will substitute detrimental for beneficial mutations.
Second, to study how accumulating mutations impact fitness and to learn how to correct this, I will engineer
plants with known mutator and anti-mutator alleles. These alleles, associated with the DNA repair machinery
and cancer susceptibility, can increase or decrease the mutation rate in A. thaliana, helping us explore mutation
accumulations up to lethal levels in many mammals. Overall, my res...

## Key facts

- **NIH application ID:** 11123794
- **Project number:** 7DP5OD029506-05
- **Recipient organization:** UNIVERSITY OF CALIFORNIA BERKELEY
- **Principal Investigator:** Moises Exposito-Alonso
- **Activity code:** DP5 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $401,250
- **Award type:** 7
- **Project period:** 2020-09-10 → 2026-08-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/11123794

## Citation

> US National Institutes of Health, RePORTER application 11123794, Predicting and controlling polygenic health traits using probabilistic models and evolution-inspired gene editing (7DP5OD029506-05). Retrieved via AI Analytics 2026-05-21 from https://api.ai-analytics.org/grant/nih/11123794. Licensed CC0.

---

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