GOALS The goals of the class are to introduce statistical
concepts and tools necessary to interpret and critically evaluate the
literature on gene expression array data, and to survey open statistical
challenges in the analysis of gene expression. The format is a hybrid
between a journal club and a traditional class. The list of articles
includes some of the most prominent array analyses from the biological
and medical literature, as well as a selection of papers from the
statistics and computer science literature. Advanced statistical
material will be presented at an intuitive level. However, knowledge of
basic probability and familiarity with regression are a
prerequisite.
COURSE SYLLABUS AND MATERIALS
Please click on the painting to access the detailed syllabus and course
materials.
A detail from Paul Cézanne's "Pommes et
biscuits" (normal and overexpressed apples), 1879-1882 from the
Musée de l'Orangerie, Paris. Cézanne was a precursor of expressionism in
more ways than it is sometimes realized....
MATCHMAKING
Thanks to all students who sent their information. Here are the lists for
possible matchmaking: Closer to biology: Liying Zhang
lzhang@jhsph.edu, Xiao Liu xiaoliu@mail.jhmi.edu, Thomas Cappola
tcappola@jhmi.edu,
Arne IJpma arne@mail.jhmi.edu
Closer to stats: Fang-Chi Hsu fhsu@jhsph.edu, Dongmei Liu
, Zhijin Wu zwu@jhsph.edu, Hongling Zhou hzhou@jhsph.edu
Somewhere in between: Patty Gravitt gravittp@mail.nih.gov
plus some others as yet unclassified....
TOPICS The syllabus is a broad overview of all steps of
microarray data production, management and analysis, with deeper
digressions into clustering, classification, and statistical pattern
recognition. Topics include: - Background on gene expression
- Expression array technologies
- Basic array bioinformatics
-
Algorithms for image analysis
- Normalization and bias adjustments;
floor and ceiling effects
- Estimating ratios: the small denominator
trap
- Importance of replication
- Clustering
- Selecting of
genes for further study
- Dimension reduction and orthogonalization;
SVD; Principal components.
- Binary regression
- Recursive
partitioning; CART
- Partial least squares
LINKS
|