TY - JOUR
T1 - Bayesian Detection of Abnormal Asynchrony of Division between Sister Cells in Mutant Caenorhabditis elegans Embryos
AU - Liang, Wei
AU - Yang, Yuxiao
AU - Fang, Yusi
AU - Zhao, Zhongying
AU - Hu, Jie
N1 - Funding Information:
The authors would like to thank Dr. Da Zhou and Dr. Jing Cheng for their support and discussion of this work. This work is supported by the National Natural Science Foundation of China, grant no.: 11601453, 11701484; the Natural Science Foundation of Fujian Province of China, grant no.: 2017J05013; and the Fundamental Research Funds for the Central Universities in China, grant no.: 20720160004.
PY - 2019/5
Y1 - 2019/5
N2 - Cell division timing is critical for cell fate specification and morphogenesis during embryogenesis, but how division timings are regulated among cells during development is poorly understood. In this article, we focus on the comparison of asynchrony of division, that is, difference of lifetime, between sister cells (ADS) among wild-type and mutant individuals of Caenorhabditis elegans. On the one hand, due to extreme imbalance between wild-type individuals and mutant-type samples, direct comparison of two distributions of ADS between wild type and mutant type is not feasible. On the other hand, we originally found that the ADS is correlated with the lifespan of the corresponding mother cell in wild type. Hence, a semiparametric Bayesian quantile regression method where lifetime of the mother cell is taken as covariate is developed to estimate the 95% confidence curve of ADS in wild type and then ADS of mutant type is classified as abnormal if outside the corresponding confidence interval. A high accuracy of our method is demonstrated by a large-scale simulation study. Real data analysis shows that ADS is related to gene function and expression quantitatively.
AB - Cell division timing is critical for cell fate specification and morphogenesis during embryogenesis, but how division timings are regulated among cells during development is poorly understood. In this article, we focus on the comparison of asynchrony of division, that is, difference of lifetime, between sister cells (ADS) among wild-type and mutant individuals of Caenorhabditis elegans. On the one hand, due to extreme imbalance between wild-type individuals and mutant-type samples, direct comparison of two distributions of ADS between wild type and mutant type is not feasible. On the other hand, we originally found that the ADS is correlated with the lifespan of the corresponding mother cell in wild type. Hence, a semiparametric Bayesian quantile regression method where lifetime of the mother cell is taken as covariate is developed to estimate the 95% confidence curve of ADS in wild type and then ADS of mutant type is classified as abnormal if outside the corresponding confidence interval. A high accuracy of our method is demonstrated by a large-scale simulation study. Real data analysis shows that ADS is related to gene function and expression quantitatively.
KW - ADS
KW - Bayesian statistics
KW - quantile regression
KW - semiparametric model
UR - http://www.scopus.com/inward/record.url?scp=85065587790&partnerID=8YFLogxK
U2 - 10.1089/cmb.2018.0246
DO - 10.1089/cmb.2018.0246
M3 - Journal article
C2 - 30964328
AN - SCOPUS:85065587790
SN - 1066-5277
VL - 26
SP - 495
EP - 505
JO - Journal of Computational Biology
JF - Journal of Computational Biology
IS - 5
ER -