Citations

Abstract With the rapid advancements in large language model technology and the emergence of bioinformatics‐specific language models (BioLMs), there is a growing need for a comprehensive analysis of the current landscape, computational characteristics, and diverse applications. This survey aims to address this need by providing a thorough review of BioLMs, focusing on their evolution, classification, and distinguishing features, alongside a detailed examination of training methodologies, datasets, and evaluation frameworks. We explore the wide‐ranging applications of BioLMs in critical areas such as disease diagnosis, drug discovery, and vaccine development, highlighting their impact and transformative potential in bioinformatics. We identify key challenges and limitations inherent in BioLMs, including data privacy and security concerns, interpretability issues, biases in training data and model outputs, and domain adaptation complexities. Finally, we highlight emerging trends and future directions, offering valuable insights to guide researchers and clinicians toward advancing BioLMs for increasingly sophisticated biological and clinical applications.

Abstract Germ line mutations in several genes (BRCA1, BRCA2, and CHEK2) whose products are involved in the DNA damage–signaling pathway have been implicated in prostate cancer risk. To identify additional genes in this pathway that might confer susceptibility to this cancer, we analyzed a recently identified DNA damage–response gene, p53AIP1 (a gene encoding for p53-regulated apoptosis-inducing protein 1), for genetic variants in prostate cancer. Five novel germ line variants were identified. The two truncating variants (Ser32Stop and Arg21insG) were found in 3% (4 of 132) of unselected prostate tumor samples. Genotyping of the two variants in an additional 393 men with sporadic prostate cancer showed a frequency of 3.1% (12 of 393) in contrast to 0.6% (2 of 327) in 327 unaffected men (Fisher's exact test, P = 0.018), with an odds ratio (OR) of 5.1 [95% confidence interval (95% CI), 1.1-23.0]. In addition, two of six tumors carrying the truncating variants were associated with loss of heterozygosity of the wild-type alleles, suggesting that p53AIP1 may act as a tumor suppressor. We also showed that the truncated p53AIP1 was unable to induce apoptosis and suppress cell growth in HeLa and COS-7 cells. These results suggest that loss-of-function variants in p53AIP1 associated with the risk of sporadic prostate cancer and further support the concept that the genetic defects in the DNA damage–response genes play an important role in the development of prostate cancer. (Cancer Res 2006; 66(21): 10302-7)