|
|
|
|
|
|
Tel Aviv University
-- Blavatnik
School of Computer Science
Seminar 0368-3178-01
Towards the Precision
Medicine Era: Computational challenges
ñîéðø: ì÷øàú òéãï äøôåàä äîåúàîú àéùéú: àúâøéí çéùåáééí
http://www.cs.tau.ac.il/~rshamir/seminar/16/precmedsem16.html
Prof. Ron Shamir
Fall 2016
Tuesdays 3-5 pm, Shenkar Physics bldg. 105
Topic: The era of personalized (or precision)
medicine is behind the corner. The combination of cheap and accessible biotechnology,
advanced computation and Big Data is expected to change the way drugs and
treatments are administered to patients: rather than one-size-fits-all, they
will be tailored to the particular properties of a group of individuals. These
properties can be based on their genomes (available via DNA deep sequencing),
their life style (known via online monitoring and wearable devices) and their
medical history (available as electronic medical records). This tailoring,
i.e., the determination of the best treatment based on the parameters, raises
major computational challenges, and we shall aim to study some of those in the
seminar.
Large projects around the
world have been initiated over the past few years towards that era. Among those
are President Obama's Precision
Medicine Initiative in the US, Genomics England 100,000
Genomes Project, and Denmark's GenomeDenmark platform.
Commercial companies like 23andme
and Regeneron
and Geisinger
Health Systems have collected genetic and clinical data from
hundreds of thousands of patients. The grand challenge is how to make the best
use of such data.
The techniques in the
papers that we shall discuss come from the areas of algorithms, statistics and machine learning. A big part of the learning effort will be to
understand the difficulties and peculiarities of the specific data types.
The seminar is open for BSc AND MSc students.
Among master students, those in the bioinformatics track will have priority.
Other interested students should inquire with the instructor.
Prerequisites:
Passing successfully the courses Statistics for
CS and Algorithms. Background in machine learning and bioinformatics is
advantage but not a must. The
basic background in biology will be given in the first meetings.
Course
material:
· Syllabus (tentative)
Plan
Note that some topics in
the syllabus have multiple papers. Only the first is listed here.
|
Date |
Speaker |
Topic |
Paper |
Method |
Presentation |
|
1/11 |
Ron Shamir |
Introduction - biology |
|||
|
8/11 |
Ron Shamir |
Introduction – precision
medicine |
|
||
|
15/11 |
Nimrod Rapoport |
Breast cancer subclasses |
Netanely (16) |
K-means, FDR, Kaplan-Meyer plots, log-rank
test, Cox uni and multivariate analysis, Wilcoxon rank
sum test |
|
|
21/11 |
Ron Shamir |
Mammaprint – Breast cancer prognosis
biomarkers |
Van't veer (02) |
Clustering, classification |
|
|
29/11 |
No meeting |
|
|
||
|
6/12 |
Gal Dinstag |
patient specific driver gene
prediction |
Nagarajan (15) |
OncoImpact , shortest paths, randomization
testing |
Lec 5 |
|
13/12 |
Ron Gal |
Pan cancer analysis |
Leiserson (15) |
Insulated heat diffusion process |
Lec 6 |
|
20/12 |
Zohar Manber |
Phenome-wide scan of gene-disease associations |
Denny (10, 13) |
Association, p-value, chi-square tests,
multiple hypothesis corrections |
Lec 7 |
|
27/12 |
Nomi Hadar |
Personalized prediction of glycemic response |
Segal (15) |
Gradient boosting regression, decision
trees, partial dependence plots |
Lec 8 |
|
3/1/17 |
Chen Arviv |
Autism classification |
Lingren (16) |
Rule-based classification, SVM, clust4ering,
dimension reduction |
|
|
|
|
Contact info: email: rshamir AT tau dot ac dot il;
phone: 640-5383; office: Schreiber 014; office hours – by appointment
picture credits:
·
http://bioinformaticsreview.com/20151005/biominer-intro/
·
Time magazine
·
https://www.linkedin.com/pulse/20140923215637-5241481-artificial-intelligence-to-deliver-personalised-medicine
·
https://www.whitehouse.gov/blog/2015/01/30/precision-medicine-initiative-data-driven-treatments-unique-your-own-body