Back

A proteogenomic approach to identifying a gene signature associated with HOXB13 G84E carrier status in prostate cancer tumours

Raspin, K.; Noor, Z.; Aref, A. T.; Marthick, J. R.; Jung, C.-h.; Donovan, S.; Williams, S.; Hains, P. G.; Robinson, P. J.; Balleine, R. L.; Zhong, Q.; Reddel, R. R.; FitzGerald, L. M.; Dickinson, J. L.

2025-07-10 genetic and genomic medicine
10.1101/2025.07.08.25331050 medRxiv
Show abstract

The rare HOXB13 G84E variant is one of few established high-risk prostate cancer variants, yet the underlying molecular mechanisms by which it drives tumour development remain poorly understood. Prostate tumours from six G84E variant carriers and 18 non-carriers (wildtype) were examined using the Human AmpliSeq Transcriptome Gene Expression Panel and mass spectrometry. We identified a 799-gene expression profile and a 75-protein expression profile in G84E positive tumours when compared to wildtype tumours. Cluster analysis revealed that the top cluster consisted of 23 proteins which mapped to several lipogenesis pathways. Integration of the transcriptome and proteome data revealed overlapping differential expression of three genes (NCLN, NFIB and VASN), two of which have an established link to lipid metabolic pathways. Examination of publicly available transcriptome data from 493 TCGA primary PrCa tumours revealed 52 tumours with a similar pattern of differential expression of these three genes. We provide the first evidence from patient samples that the G84E variant is associated with upregulation of genes in the steroid metabolic, lipid biosynthesis and fatty acid metabolic pathways. We are also the first, to our knowledge, to use a proteogenomic approach to identify key molecular changes in prostate tumours from any high-risk rare variant carriers.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.