Shirng-Wern Tsaih

Randy SmithFormer Research Scientist
Churchill Group, The Jackson Laboratory


1. Quantitative Trait Loci Mapping for Recombinant Inbred Strains
Recombinant inbred (RI) strains represent a critical tool for the genetic dissection of simple and complex traits. A common way to map quantitative trait loci (QTL) in RI lines is to include only a single measurement from each line representing the average among the replicates (a strain mean model). Several lines of data suggest that genetic interactions may be important to the sex differences observed in the traits. The common approach used in QTL mapping in RI strains is to analyze the male and female data separately. In this study, we describe and test a method of analysis, based on a linear mixed model, that accounts for the correlation structure between male and female individuals of the same RI strain in QTL mapping. We propose a genotype permutations test, which conserves the genetic structure of the RI panel to correctly control type I error. Finally, we applied the mixed model approach to map QTL in femur phenotypes in the large panel of L x S RI mouse strains, developed from inbred long-sleep (ILS) and inbred short-sleep (ISS) strains.
In collaboration with Dr. James F. Nelson's Lab at the University of Texas Health Science Center at San Antonio, we will apply this mixed model approach to identify sex or/and diet-specific QTL for body weight, fat weight and lean mass in the LxS RI strains aging study.

2. Causal Factors in Mouse Models of Aging
My preliminary analysis of lifespan data among 31 inbred mouse strains (The Jackson Shock Center) showed that traits measured at young age of mice had better correlation with median lifespan than those measured at older age. The patterns of correlation with median lifespan data also varies by sex. Body composition traits and insulin-like growth factor-I were negatively correlated with median lifespan, whereas red blood cell counts (RBC) were positively correlated with median lifespan data. Analysis of variance test reveals that both strain and age significantly impact RBC in male and female mice. Substantial between and within strain sex variability in the decline in RBC levels with age is also observed. The goal of my proposal is to identify measurable traits that can serve as reliable early indicators of longevity and age related pathologies. I will focus on applying graphical model technique utilizing multiple age-related phenotypes in the coming years to gain insights into the biological changes that occur with age.

3. Intergenerational effect of dietary restriction diet in laboratory mice
Data of people exposed to Dutch famine in utero at the end of World War II provides evidence that severe maternal calorie restriction at different stage of pregnancy can program insulin resistance and type 2 diabetes in the offspring. Dietary restriction (DR) is the most robust and reliable means of extending life in rodents. Therefore, it is important to evaluate the intergenerational effect of DR. My project will employ high-density microarrays to define transcriptional patterns across multiple tissues from mice carrying DR history from their parents.

4. Computational and statistical core at The Jackson Aging Center
I support the data and workflow management of the Phenome Project at The Jackson Shock Center in the early phase and to carry out the sophisticated statistical and genetic analysis in the latter phase. The analyses include survival analysis, multivariate analysis, haplotype association mapping and structural equation modeling to identify the phenotypes and QTL associated with aging.

5. Other QTL mapping projects

  • Metabolic syndrome QTL Project
    In collaboration with Dr. Fumihiro Sugiyama at the University of Tsukuba in Japan, I am working to identify QTL influence metabolic syndrome in HZO/H1LtJ×C3H/HeJ mouse intercrosses. Structural equation modeling will be applied to sort out the inter-relationship among correlated phenotypes and QTL.

  • Hematology QTL Project
    In collaboration with the Jackson lab investigator Dr. Luanne Peters, I am working to identify QTL influencing baseline erythropoietic traits and baseline WBC counts in eleven mouse crosses. We are currently working to narrow the critical chromosome intervals to facilitate future candidate gene analyses using haplotype analysis and combining cross analysis methods.
  • Chronic kidney disease QTL Project
    In collaboration with Dr. Keith DiPertrillo in Novartis Institutes for BioMedical Research, we are working to identify genetic factors that influence chronic kidney disease in nine mouse crosses. The goal is to narrow the critical chromosome intervals to facilitate future candidate gene analyses using techniques such as combining cross analysis and haplotype analysis.

  • Hypertension QTL Project
    In collaboration with Dr. Keith DiPertrillo in Novartis Institutes for BioMedical Research, we are working to identify genetic factors regulate hypertension in eight mouse crosses to narrow the critical chromosome intervals to facilitate future candidate gene analyses.

  • Bone QTL Recalculation Project
    There are many inconsistencies with regards to marker order and relative positions in QTL publications. In collaboration with the Jackson lab investigator Dr. Beverly Paigen, we are in the process of recalculating the peak position and confidence intervals for BMD QTLs using raw data with updated genetic map from eleven mouse crosses

Mixed Model Analysis of Quantitative Traits in Recombinant Inbred Mouse Strain Panels (PDF)
Shirng-Wern Tsaih
, Sarah R Langley, Wesley Beamer, Beth Bennett, Gary A Churchill
7th Annual Meeting of the Complex Trait Consortium in Montreal Canada, May-June 2008. Poster

Haplotype Association Mapping in mice (PDF)
Shirng-Wern Tsaih
Genomic and Proteomic Approaches to Complex Heart, Lung, Blood, & Sleep Disorders, September 10, 2008. Talk