Zhengdeng Lei, PhD

Zhengdeng Lei, PhD

2009 - Present Research Fellow at Duke-NUS, Singapore
2007 - 2009 High Throughput Computational Analyst, Memorial Sloan-Kettering Cancer Center, New York
2003 - 2007 PhD, Bioinformatics, University of Illinois at Chicago

Wednesday, June 6, 2012

Principal Investigator@Ontario Institute for Cancer Research

http://bioinformatics.ca/resources/jobs/principal-investigator


Institution/Company: 
Ontario Institute for Cancer Research
Location: 
Downtown Toronto
Job Description: 
Position: Principal Investigator
Site: MaRS Centre, Toronto
Department: Informatics & Bio-computing
Reports To: Director, Informatics & Bio-computing Platform
Salary: Commensurate with level of experience
Hours: 35 Hrs/week
Status: Full-time, Permanent
The Ontario Institute for Cancer Research (OICR) is seeking Junior, Intermediate and Senior Principal Investigators (PIs) in Bioinformatics, Computational Biology and Biostatistics to undertake world-class computational research in a wide range of research areas, including any of the following: (1) discovery of key genetic alterations in the initiation or progression of cancer; (2) identification of biomarkers indicative of tumour subtypes or predictive of response to targeted therapy; (3) modeling of regulatory networks relevant to disease pathways; (4) analysis of genetic and environmental risk factors for cancer in patient populations; (5) use of machine learning and/or biostatistical approaches to develop novel algorithms and computational techniques for genomic and/or epigenomic data; (6) development of interoperability standards for exchanging and collaboratively annotating genome-scale data sets; or (7) development of software engineering techniques for managing and manipulating genome-scale datasets and complex analytic workflows. We expect to appoint up to five PIs over the period 2012-2014.
PIs will be expected to mentor trainees, and to build collaborations both within and outside the OICR community. In addition to base funds provided by the Institute to support the PI's salary and personnel, PIs are expected to raise additional research funds from external competitive granting agencies. The OICR will assist PIs in obtaining faculty appointments at the University of Toronto or another affiliated academic institution.
QUALIFICATIONS
• An MD or PhD with a proven track record in computational biology, bioinformatics, or biostatistics;
• For new PIs, a record of independent research and either first-author peer reviewed publications or the publication of software, databases or other significant community resources.
• For senior and intermediate-level PIs, international recognition and a strong publication record of relevance, proven leadership and management experience including the building of strong research teams, as well as a strong record of mentorship and/or teaching;
• Eligible to hold the rank of assistant, associate or full professor at an Ontario university;
• Excellent communication and presentation skills.
OICR is an innovative cancer research institute located in the MaRS Centre in the Discovery District in downtown Toronto. OICR is addressing significant challenges in cancer research with multi-disciplinary, multi-institutional teams. New discoveries to prevent, detect and treat cancer will be moved from the bench to practical applications in patients. The OICR team is growing quickly. We are innovative, dedicated professionals who bring expertise to each of our roles. We are looking for individuals interested in being part of a culture of excellence that will result in Ontario being recognized internationally as a leading jurisdiction for cancer research.
Launched in December 2005, OICR is an independent institute funded by the Government of Ontario through the Ministry of Economic Development and Innovation.
For more information about OICR, please visit the website at www.oicr.on.ca.
POSTED DATE: June 1, 2012
CLOSING DATE: Posted until filled
Interested candidates may apply here
https://www.recruitingsite.com/csbsites/oicr/JobDescription.asp?JobNumber=675388
OICR has a diverse workforce and is an equal opportunity employer.
The Ontario Institute for Cancer Research thanks all applicants. However, only those under consideration will be contacted. Candidates will be expected to provide their current employer as a reference.
Resume Format: If you elect to apply, you will need a text or HTML version of your resume so that you can cut and paste it into the application box provided. Before you submit the completed application, you will be asked to attach one or two files to your application. Please attach your resume as a .doc file.

try

1. use microdissected BC-class to predict NCI60/GEMINI
ANSWER: looks like not good.

2. use macrodissected BC-class to predict BC57 with some dropped samples (to be population balance)
Answer: tried, but not good result, so it is not because of population bias.

Tuesday, June 5, 2012

Normalization before NTP

Before NTP, is it better to do standardization on gene(row) then on array(column) to avoid population bias.
Or may median polish (an iterative method)?

ANSWER: It is NOT GOOD to do standardization on gene(row) then on array(column) to avoid population bias.


################################################
# standardization by row and column, respetively
#By row (gene)
std.data.by.row <- t(scale(t(data.filtered), scale=T))
#By column (array)
std.data <- scale(std.data.by.row, scale=T)





Macrodissection versus microdissection of rectal carcinoma: minor influence of stroma cells to tumor cell gene expression profiles




As microdissection yielded low tissue and RNA quantities, extra rounds of mRNA amplification were necessary to obtain sufficient RNA for microarray experiments. These second rounds of amplification influenced the gene expression profiles. Moreover, the presence of stroma cells in macrodissected samples had a minor contribution to the tumor cell gene expression profiles, which can be explained by the observation that more RNA is extracted from tumor epithelial cells than from stroma.

Monday, June 4, 2012

two classifiers for cancer

In practice, we need two classifiers, one from microdissected tumor samples, the other from macrodissected tumor samples. Hence we can predicted microdissected samples by microdissected classifier, and macrodissected samples by macrodissected classifier.


Q: Is it not good to predicted microdissected samples by macrodissected classifier?
Answer: NOT good, confirmed.


Q: Really? It could also be due to population bias between training set and test set?
Answer: Not because of population bias


Q: it may be okay to predict macrodissected samples by microdissected classifier.?
Answer: NO, the other way around isn't work either.


 check with breast cancer datasets.

duke-nus center for cb

65165240
http://www.ezlink.com.sg/top-up/ez-reload-apply2.php?card=yes

Saturday, June 2, 2012

Macrodissection vs microdissection


Macrodissection versus microdissection of rectal carcinoma: minor influence of stroma cells to tumor cell gene expression profiles