Fig. 1: Causal diagrams representing latent factor models. | Communications Biology

Fig. 1: Causal diagrams representing latent factor models.

From: Capturing additional genetic risk from family history for improved polygenic risk prediction

Fig. 1

a For the parents and the children separately, a continuous trait Y is determined by a polygenic component P captured by a polygenic risk score, and a latent genetic component G independent of P. b A binary trait Z is determined by the underlying genetic liability H. Analogous to Y in a, H is jointly determined by P and G. Prediction combining a polygenic risk score and parental information is equivalent to inference of the distribution of Y in a and H in b for the children. The two genetic components, P and G, are assumed to be independently passed on from the parents to the offspring. Trait heritability is assumed to be constant across different generations.

Back to article page