News & Events

“Paternal Valproate Exposure and Offspring Neurodevelopmental Outcomes” published in Neurology

In collaboration with CGMH and NHRI, this study looks at whether paternal exposure to valproate (a common and effective drug for epilepsy and bipolar disorder etc.) during spermatogenesis would increase the risk of NDD in their children, using the NHIRD database in Taiwan.

This has been a contentious issue since 2023: an initial report by IQVIA using a Nordic registry found significantly elevated risk, prompting a warning in the UK MHRA to advise against valproate use for male patients less than 55 years of age. However, several follow-up studies—now including our non-Nordic data, with several complementary designs to mitigate confounding—reported otherwise (i.e., null association), calling for a reconsideration of valproate regulation in men of reproductive age.

Reference:
Feng, Y.-C. A., Lin, M.-C., Tseng, C.-H., Wu, C.-S., Tsai, M.-H.*, Wang, S.-H.* (2026). Paternal valproate exposure and offspring neurodevelopmental outcomes. Neurology, 107(6), e218375.

Editorial Comment:
Voinescu, P. E. (2026). Paternal valproate and offspring neurodevelopment: A non-Nordic perspective on an unsettled question. Neurology, 107(6), e218516.

“Variability in epilepsy polygenic risk prediction across Taiwanese population and clinical cohorts” published in Epilepsia

Our study examined whether polygenic risk scores—measures that combine the effects of many genetic variants—can help identify people at higher risk of epilepsy in more than 600,000 Taiwanese individuals. We found that these genetic scores performed best in carefully characterized clinical samples and were generally more informative for people with earlier-onset epilepsy, particularly genetic generalized epilepsy (GGE). However, their predictive power varied substantially depending on the type of epilepsy and how cases were identified, highlighting the need for caution before using them in clinical practice. Our findings also underscore the importance of improving genetic studies in diverse populations and combining genetic information with richer clinical data to make precision medicine for epilepsy more accurate and equitable.

Reference:
Ke, Y.-S., Tsai, M.-H., Ho, C.-R., Chou, C.-Y., Chen, H.-H., Stevelink, R., …, Sung, S.-F., Lee, M.-C., Chen, J.-T., & Feng, Y.-C. A.* (2026). Variability in epilepsy polygenic risk prediction across Taiwanese population and clinical cohorts. Epilepsia, epi.70353.

Proud of Yi-Syuan Ke for delivering a talk on our work dissecting depression heterogeneity with Transformer architecture

A four-day scientific feast, packed with new advances pushing the frontiers of human genetics. Particularly excited about the growing possibilities of multi-omic data integration, discovery, and new methodologies for deciphering disease mechanisms.

Wishing Ke all the best as she starts her PhD journey at Aarhus University in Denmark!

“Genetic regulation of methylation across East Asian and European populations” published in Nature Communications

台大公衛學院健統所馮嬿臻老師團隊與國衛院、美國 Broad Institute 合作，發表迄今最大規模的東亞族群 DNA 甲基化遺傳調控研究，刻畫其與複雜疾病風險的關聯。

DNA 甲基化是基因體上重要的化學修飾，可調節基因表現並影響疾病風險，其本身亦受遺傳因素所調控；然而，目前相關研究以歐洲人群為主，東亞族群資料相對匱乏。本研究整合共 7,619 人的漢人族群樣本，鑑定出數十萬個受基因調控的甲基化位點，其中 28,978 個為新興發現。東亞與歐洲族群的甲基化調控機制大多共享，而東亞特有訊號則多來自在歐洲族群中較為少見的變異。此外，本研究利用孟德爾隨機化之因果分析，將遺傳變異、甲基化、基因表現與疾病風險串連，為心血管疾病、糖尿病等複雜疾病的潛在分子機轉提供新線索。研究團隊並公開釋出此東亞 mQTL 資源，為後續族群遺傳研究奠定重要基礎。

Reference:
Liu, R., Chen, T.-T., Xia, Y., Lin, S.-C., Ge, T., Chen, C.-Y.*, Feng, Y.-C. A.*, Huang, H.*, & Lin, Y.-F.* (2026). Genetic regulation of methylation across East Asian and European populations. Nature Communications, 17(1), 2616.

“You’re not just part of the sequence; You complete it.”

Attending ASHG this year as a group of four—proud of all the grad students for their hard work and growth. Excited for the science, the conversations, and what’s ahead.

“Polygenic dissection of treatment-resistant depression with proxy phenotypes in the UK Biobank” published in Journal of Affective Disorders

For about one in three people with major depressive disorder (MDD), standard antidepressant treatments don’t work. This form, called treatment resistant depression (TRD), leaves patients struggling despite trying at least two different medications.

Our team asked: could genetics help explain who develops TRD—and even guide more personalized treatments?

Using primary-care electronic health records from 230,000 people in the UK Biobank, we identified those likely to have TRD based on their prescription history. Compared to others with depression, these individuals often had an earlier onset of depression, lower employment and education levels, and more frequent hospitalizations.

The genetic findings were clear: TRD showed a much stronger inherited component (20–30% of risk) than depression overall (~8%). People with higher polygenic risk scores (PRS) for MDD, ADHD, and bipolar disorder were more likely to switch medications or use mood stabilizers, while those with a higher genetic risk for schizophrenia specifically predicted the use of antipsychotics. Higher bipolar PRS also increased the likelihood of using electroconvulsive therapy (ECT), lithium, or valproate by 1.27–1.80 fold.

These results support that TRD has a non-trivial genetic basis and highlight the potential of PRS to uncover pharmacogenomic effects that may aid treatment options for managing TRD.

Reference:
Wang, L.-H., Shih, M.-Y., Lin, Y.-F., Kuo, P.-H., & Feng, Y.-C. A.* (2025). Polygenic dissection of treatment-resistant depression with proxy phenotypes in the UK Biobank. Journal of Affective Disorders, 381, 350–359.

Shih-Hsiang (Shaun) Luo received the highest honor at NTU for his bachelor thesis, a multifactorial risk assessment of depression in the Taiwanese population – Congratulations!