Mapping the behavioral spectrum of autism in mice

In this thesis, we describe several advancements in the analysis of behavioral
phenotyping of autism mouse models. We develop novel methods to both
standardize and automate the analysis of stereotyped behaviors in chapter 2
and motor function in chapter 3, to expand our understanding of the meaning
behind some of the behavioral tests that were used. We also use these
tools to increase our understanding of the behavioral spectrum of autism,
originating from either primary immunodeficiency disorders in chapter 4 or
genetic mutations in autism risk genes and disorders commonly associated
with autism in chapter 5. We combine the recent advances in automated
classification of both discrete behaviors and underlying behavioral syllables
in chapter 5 to develop a new dimensionality reduction method that can be
applied to a range of behavioral data, to show meaningful lower-dimensional
behavioral dynamics capturing sex- and environment-dependent variability
of behavioral phenotypes in autism mouse models. We aim to quantify some
of the behavioral variability shown in mouse models of autism, to encourage
standardization in the field of behavioral phenotyping of autism mouse models.
We further build on this in chapter 6, where we summarize the behavioral
and physiological phenotypes which have been described in cerebellar-specific
genetic and circuit perturbations of autism risk factors, to gain insight into the
range of observable characteristics resulting from perturbations localized to a
brain area strongly associated with autism.