|
Statistics
Statistics is the science of data: designing how to collect data, exploring data, and drawing conclusions from data.
All data reflect truth with systematic and random error. When data are collected by the science, they will properly represent the truth.
And when data are explored by the science, they will be
interpreted correctly.
Figures often beguile me, particularly when I have the arranging of them myself; in which case the remark
attributed to Disreaeli would often apply with justice and force: "There are three kinds of lies: lies, damned lies,
and statistics." - Autobiography, Mark Twain
... so we need statisticians.
|
Bioinformatics
In addition to complete genome of various species, databases with genomic scale annotations
are being built. Biologists and clinicians now often measure genome-wide data using various types of microarrays.
Bioinformatics,
or statistical genomics, is a research field about methodologies for explorative analysis of genomic data. Often computational or statistical methodologies predict biological
findings and
further
experiments validate them.
In one collaboration, we identified molecular markers for differentiation of astrocytic cells in mice brain (PNAS, 2004). Now we are working on molecular studies of in vivo mouse
brain
development, and human brain meningioma. Statistical methodologies that I tackled in genomic studies are: resampling-based clustering algorithm, supervised feature selection, gene-set
analysis, and cis-regulatory analysis.
|
People
Richard Yanicky (2007.3~)
Richard is a part-time master student and works full-time at Pfizer. We are working on analysis of human meningioma with comparative genomic
hybridization (CGH) arrays and expression arrays.
Ruijuan Guo (2007.5~)
Ruijuan is a master student. We are working on the low-level analysis of illumina expression arrays.
Mei Jin (2007.11~)
Mei is a new PhD student. We are working on general topics of bioinformatics.
Alumni
Qing Hu (2007.1~2007.05) Qing is now at Esurance after earning her master's degree. We worked
on adapting and applying
L1 regularization procedure to select covariates and shrink estimates in regression analysis.
|