At the center of her recognition is Geneformer, described by Gladstone as the world’s first foundational AI model created to predict how changing gene activity impacts individual human cells. According to the announcement, Geneformer has already identified promising new drug targets for heart disease. Building on that work, her lab recently introduced MaxToki, a temporal AI model capable of predicting cellular changes associated with aging and identifying methods to slow that process. Her lab’s models are designed to predict how gene networks regulate development, aging, and disease progression, work that the institute describes as having immediate applications for human health.

“Receiving the Seldin-Smith Award is an incredible honor and a wonderful recognition of the promise of our science to make a clinical impact,” Theodoris said in the announcement. “My team is deeply committed to developing AI models to uncover fundamental principles of gene network biology and accelerate the discovery of new, transformative therapies for complex diseases.” Benoit Bruneau, PhD, director of the Gladstone Institute of Cardiovascular Disease, praised her work at the intersection of AI and biology, saying that by developing generative models to map complex cellular dynamics, her research is opening new avenues to overcome disease and improve patient outcomes.

The award carries a historical resonance for Gladstone. Established in 2014, the Seldin-Smith Award honors Donald W. Seldin, MD, and Lloyd H. “Holly” Smith Jr., MD, two celebrated ASCI members who passed away in 2018 and are remembered for their research contributions, institutional leadership, and mentorship of generations of physician-scientists. Smith, a longtime Chair of Medicine at UC San Francisco, played a central role in UCSF’s rise to international prominence and in the founding of the Gladstone Institutes in 1979. Seldin played a similar role in the emergence of the University of Texas Southwestern Medical Center in Dallas as a top biomedical center.

Deepak Srivastava, MD, president of Gladstone, who mentored Theodoris during her graduate studies as an MD-PhD student at UCSF a decade ago, called her an exemplar of the boundary-pushing creativity and clinical insight the award was designed to celebrate. He noted that it is profoundly meaningful for the community to see her receive an honor bearing the name of Holly Smith, whose vision shaped Gladstone’s history. Along with the recognition, the award includes dedicated mentorship from members of the award selection and advisory committees. Theodoris will also be inducted into the ASCI and deliver a keynote research presentation at the combined annual meeting of the ASCI, the Association of American Physicians, and the American Physician Scientists Association in April 2027.

Subject of Research: Award recognition of physician-scientist Christina Theodoris for artificial intelligence models of gene network biology

Article Title: Christina Theodoris to receive ASCI’s 2026 Seldin-Smith Award for Pioneering Research

Article References: Christina Theodoris to receive ASCI’s 2026 Seldin-Smith Award for Pioneering Research. Original publication (Bibliographic details unavailable.)

Image Credits: AI Generated

DOI: Not provided

Keywords: Christina Theodoris, Seldin-Smith Award, ASCI, Gladstone Institutes, artificial intelligence, Geneformer, gene networks, cardiovascular disease, aging, physician-scientist, UC San Francisco, Christina

The study included 38 people with essential tremor and 37 healthy controls. Investigators tested blood-gene expression differences with DESeq2, adjusting for age, sex, RNA-sequencing technical quality and estimated immune-cell composition. After correction for multiple comparisons, four genes were significantly elevated in the essential tremor group: CXCL5, HSPB1, CLEC7A and PRDX6. These case-control effects were sparse, but they remained significant across sensitivity analyses. CLEC7A also showed nominal concordance in an independent dataset of RNA-sequenced cerebellar tissue from people with essential tremor. More broadly, 59.9% of genes nominally associated with essential tremor in blood had effect directions that matched those in the cerebellum, a concordance rate described as exceeding chance expectation. This cross-tissue agreement suggests that some blood signals may correspond to biological changes measurable in brain tissue, although it does not demonstrate that blood expression can be used to diagnose the condition.

The strongest results emerged when researchers examined differences within the essential tremor group rather than simply comparing affected participants with controls. Covariate-adjusted linear models linked expression of 40 genes to scores from the Montreal Cognitive Assessment, including GNLY, ZAP70 and KLRD1. Gene-expression patterns were also associated with regional brain measurements: 253 genes correlated with right pericalcarine cortical thickness and 46 with the volume of the anterior corpus callosum. The analysis did not find significant case-control differences in the study’s primary global neuroimaging measures. The contrast is important because it indicates that peripheral blood expression carried a limited signal for diagnosis but may contain information about the varied cognitive and neuroanatomical profiles found within essential tremor.

Weighted gene co-expression network analysis identified coordinated groups of genes whose activity tracked those clinical and imaging characteristics. A cytotoxic immune module was strongly associated with lower MoCA scores, with a correlation of r = −0.76 and an adjusted p-value of 0.0002. A second module, containing genes related to mitochondrial expression and translation, was associated with right pericalcarine cortical thickness (r = −0.65, adjusted p = 0.01). An erythroid and vascular transport module correlated with anterior corpus callosum volume (r = 0.70, adjusted p = 0.0046). These are associations observed in a relatively small observational cohort, not evidence that changes in those pathways cause cognitive impairment or structural brain differences. The network results nevertheless point toward immune, metabolic and vascular processes that may help explain phenotypic heterogeneity in essential tremor and justify testing whether the same relationships appear in independent and longitudinal samples.

Subject of Research: Blood transcriptome associations with clinical and neuroimaging variation in essential tremor

Article Title: Blood transcriptional variation is associated with clinical and neuroimaging phenotypes in essential tremor

Article References: Welton, T., Teo, T. W. J., Ng, E. Y., Sun, Q., Lee, B., Saffari, E., Chan, L. L., Tan, L. C. S., & Tan, E.-K. (2026). Blood transcriptional variation is associated with clinical and neuroimaging phenotypes in essential tremor. Scientific Reports. https://doi.org/10.1038/s41598-026-76750-z

Image Credits: AI Generated

DOI: 10.1038/s41598-026-76750-z

Keywords: essential tremor, blood transcriptomics, RNA sequencing, brain MRI, cognition, immune pathways, gene networks, cerebellar tissue, biomarkers, Blood, transcriptional, variation