An AI-enabled framework clarified how hidden changes in the RHD gene may influence hemolytic disease of the fetus and newborn (HDFN), offering patients and clinicians a more precise way to understand risk, according to results of a study that will be presented at the 67th ASH Annual Meeting and Exposition held Dec. 6 to 9 in Orlando, FL.
For families affected by HDFN, this work suggests future blood typing could become more accurate, which may ultimately improve monitoring and treatment decisions.
The investigators combined traditional molecular studies with AlphaGenome, an AI system launched in June 2025. AlphaGenome narrowed the complex landscape of possible RhD regulatory sites to small, highly active regions that appear to control antigen levels.
“We developed experimental systems and pioneered the use of AlphaGenome for studying RhD noncoding blood groupBlood group A classification of blood based on inherited markers found on red blood cells. The ABO and Rh blood groups are most common in HDFN. variants,” explained this study’s authors. They continued, “AlphaGenome’s ability to refine regulatory region predictions significantly enhanced research efficiency, providing an innovative tool for non-coding variant analysis.”
These areas included Chr1:25272393–25272547, which contains the promoter and 5′UTR, as well as intron–exon junctions. Across these sites, the system predicted 217 possible variants. Twenty six were extremely rare in gnomAD at less than 0.001 percent, including known variants tied to altered RhD expression such as c.1110A>C and c.1-83C>T.
Read more about the prognosis of HDFN
To understand how these changes could behave in real cells, researchers built erythroid cell models using K562 and HuDEP2 lines. They measured RhD mRNA levels, analyzed full-length and alternatively spliced transcripts, and verified protein expression by flow cytometry. K562 cells expressed strong RhD and RhCE, while HuDEP2 cells showed high RhD with low RhCE, creating a useful contrast for exploring how different versions of the gene may function.
Whole-genome sequencing of K562 cells identified 84 variants across the RhD region, though none were clinically significant. Still, the combination of sequencing, transcript analysis, and protein assays offered a solid template for confirming which noncoding variants might matter for patients. Importantly, AlphaGenome’s predictions aligned well with histone-based mapping, reinforcing confidence in the flagged regulatory sites.
For patients, the potential impact lies in more informative RHD typing. As CRISPR-based functional testing progresses, clinicians may gain tools to detect subtle genetic patterns that change how much RhD antigen is produced. Better resolution of these differences could improve risk assessment for HDFN, support safer pregnancy management, and enhance transfusion accuracy.
Sign up here to get the latest news, perspectives, and information about HDFN sent directly to your inbox. Registration is free and only takes a minute.
