Search papers, labs, and topics across Lattice.
This paper introduces the Temporal Heterogeneous Biomedical Knowledge Graph (THBKG), which captures and predicts therapeutic target-disease linkages over time, addressing a critical gap in current biomedical knowledge graphs. By incorporating historical evidence profiles for target-disease pairs at the time of clinical decisions, THBKG enables more accurate predictions of which programs will advance from Phase II to Phase III. The results show that graph propagation over THBKG significantly outperforms traditional direct-evidence references, particularly for pairs lacking direct evidence, achieving a relative success rate of 4.3-4.5 for the top ten pairs in each therapeutic area.
Graph propagation over a temporal knowledge graph can predict clinical advancement with unprecedented accuracy, especially for cases lacking direct evidence.
Inadequate target--disease linkage accounts for 40--50\% of Phase~II efficacy failures, so anticipating which programmes will advance would let sponsors back the hypotheses most likely to reach patients. What a programme can be judged on is the evidence that supported its linkage \emph{when it entered the clinic}. No existing biomedical knowledge graph allows that evidence profile to be assembled as of a past date. We present the Temporal Heterogeneous Biomedical Knowledge Graph (THBKG), which describes and predicts therapeutic target--disease links through time: 110,396 entities and 11.1M edges across nineteen relation types, each edge carrying the year its evidence changed, so a pair's profile can be recovered as it stood when its own decision fell due. On this graph we define a decision-aligned benchmark that predicts, for a target--disease pair entering Phase~II, whether it advances to Phase~III on evidence datable before that decision. Graph propagation over the THBKG outranks every direct-evidence reference scored under the same decision-aligned protocol, reaching a relative success of 4.3--4.5 at the top ten pairs per therapeutic area. The gain concentrates on the 72.8\% of pairs with no direct target--disease evidence at their decision point, where a direct-edge model has nothing to read: the encoders still rank five- to sixfold above chance, recovering the signal by propagating over the intervening biology. Adapting a path-based explainer to the decision-time subgraph decomposes each prediction into the evidence landscape behind the hypothesis for explainable prediction. We release the THBKG as a continually updated substrate for studying therapeutic target hypotheses by retrospective validation.
