1. What contributions have the authors mentioned in the paper "Efficient maximum likelihood estimation for pedigree data with the sum-product algorithm" ?
The authors analyze data sets consisting of pedigrees with age at onset of colorectal cancer ( CRC ) as phenotype.. The authors aimed to compute the probability of a family possessing this risk factor as well as the hazard rate increase for these risk factor carriers.. The authors propose an improved EM algorithm by applying factor graphs and the sum-product algorithm in the E-step.. Their algorithm is as precise as a direct likelihood maximization in a simulation study and a real family study on CRC risk.. The authors introduce a flexible and runReceived: December 13, 2016 Accepted: April 1, 2017 Published online: July 21, 2017 Alexander Engelhardt Institute for Medical Informatics, Biometry and Epidemiology Ludwig Maximilian University Marchioninistrasse 15, DE–81377 Munich ( Germany ) E-Mail engelhardt @ ibe. med. uni-muenchen. de © 2017 S. Karger AG, Basel. The occurrence of familial clusters of CRC suggests the existence of a latent, inheritable risk factor.
read more
2. What future works have the authors mentioned in the paper "Efficient maximum likelihood estimation for pedigree data with the sum-product algorithm" ?
The size of the families in such studies is likely to grow in the future, with the availabil- D ow nl oa de d by: U B d er L M U M ün ch en 12 9. 18 7. 25 4. 47 - 8 /2 0/ 20 18 8 :4 2: 09 A M Efficient MLE for Pedigree Data Hum Hered 2016 ; 82:1–15 DOI: 10. 1159/000475465 11 ity of more longitudinal data.. It is possible to estimate further parameters in this model, such as the shape parameter k.. With the sum-product algorithm, the authors can now conduct large-scale simulation studies for power and sample size determination and extract further information such as bootstrap confidence intervals from the data.
read more
![Fig. 2. Message passing during the sum-product algorithm. The messages are computed iteratively. Once all but 1 incoming messages are calculated for a given node, a message is calculated and sent across the remaining edge. Note that by Kschischang et al. [11] , the actual order in which the messages are calculated is irrelevant for the result.](/figures/fig-2-message-passing-during-the-sum-product-algorithm-the-1a3lciun.webp)
