A systematic evaluation of how model architectures and training strategies impact genomics model performance is needed. To address this gap, we held a DREAM Challenge where competitors trained models on a dataset of millions of random promoter DNA sequences and corresponding expression levels, experimentally determined in yeast. For a robust evaluation of the models, we designed a comprehensive suite of benchmarks encompassing various sequence types. All top-performing models used neural networks but diverged in architectures and training strategies. To dissect how architectural and training choices impact performance, we developed the Prix Fixe framework to divide models into modular building blocks. We tested all possible combinations for the top three models, further improving their performance. The DREAM Challenge models not only achieved state-of-the-art results on our comprehensive yeast dataset but also consistently surpassed existing benchmarks on Drosophila and human genomic datasets, demonstrating the progress that can be driven by gold-standard genomics datasets.Co-authors: Byeong-Chan Kim, Juhyun Lee, Taein Kang

Objective: Hypoparathyroidism is an uncommon but significant complication of total thyroidectomy. As the incidence of thyroid cancer increases, the prevalence of hypoparathyroidism also increases. However, its effect on the risk of osteoporotic fractures remains poorly understood. We aimed to evaluate the risk of osteoporotic fractures in patients with postoperative hypoparathyroidism compared with that in a matched control group. Design: This retrospective cohort study included 417 patients with hypoparathyroidism who had undergone total thyroidectomy for thyroid cancer (TC with hypoP) and 2085 matched controls from the Korean National Health Insurance Service-National Sample Cohort. Methods: The osteoporotic fracture rates and associated risk factors were analyzed in comparison after propensity score matching. Results: During a median follow-up period of 5.07 years (interquartile range 2.19-7.38), postoperative hypoparathyroidism was not associated with a significantly greater risk of osteoporotic fractures than that in the control group (hazard ratio [HR] 0.86, 95% confidence interval [CI], 0.59-1.24, P = .406). Age-specific analysis showed a trend toward higher fracture incidence in patients aged <50 years (HR 1.31, 95% CI, 0.53-3.22, P = .558) and a decreasing trend in those aged ≥50 years (HR 0.75, 95% CI, 0.51-1.12, P = .163), although the differences were not statistically significant. No significant differences in fracture risk were observed between the TC with and without hypoP groups across all fracture sites. The frequency of osteoporotic medication use was higher in the TC with hypoP group than in the control group, particularly among women aged <50 years. Conclusions: Hypoparathyroidism after total thyroidectomy does not appear to increase the risk of osteoporotic fractures compared with that in matched controls in clinical practice.Co-authors: Eu Jeong Ku, Won Sang Yoo, Yu Been Hwang, Subin Jang, Shinje Moon, Eun Kyung Lee, Hwa Young Ahn