Work

Myocardial splice associations after aortic valve replacement in a secondary RNA sequencing analysis

Year: 2026
Type: preprint
Abstract: Background Left ventricular hypertrophy does not regress equally in all patients after aortic valve replacement. We examined whether myocardial splice differences associated with remodelling labels persisted after individual samples were excluded, and what these differences could change in reference transcripts and proteins. Results We reanalysed 41 public myocardial RNA-sequencing samples labelled as persistent hypertrophy (HYP, n = 21) or reverse remodelling (REV, n = 20). Of 156,111 candidate splice events, 70 met the full-cohort selection criteria. Twenty-two events in 19 genes were selected with unchanged direction in all 41 leave-one-out analyses. Both local splice alternatives matched reference transcripts for 16 events. A score derived separately within each training set could be calculated for every excluded sample. Its area under the receiver operating characteristic curve was 0.648, compared with 0.998 when selection and assessment used the full cohort. DGLUCY long-site inclusion averaged 0.766 in HYP and 0.605 in REV; the difference remained positive after each sample was removed. Complete reference sequence comparisons linked the 15-base splice-site difference to the five-amino-acid segment AGAYK. This sequence alternative is already recorded in UniProt. Conclusions The analysis identified internally stable myocardial splice associations and mapped selected events to reference sequences. The modest leave-one-out discrimination and unresolved correspondence between the 41 public samples and the original 40-person clinical cohort preclude a validated prognostic interpretation. Independent outcome-linked data are needed to test these associations.
Language: English
Open Access status: green
OpenAlex ID: W7234828308
Cites: 6
Cited by:
Related to: 10