Laboratory information:
Contact information:
Sampsa Hautaniemi
Biomedicum Helsinki, B524b
P.O.Box 63 (Haartmaninkatu 8)
00014 UNIVERSITY OF HELSINKI
phone +358 9 191 25419
fax +358 9 191 25610
firstname.lastname@helsinki.fi
|
 |
Research projects
Cancer genetic data mining
Genetic variance is a major factor in several diseases, such as cancer.
One of the most frequent type of genetic variation is single nucleotide
polymorphism (SNP), i.e., a change of one nucleotide to another. There are
more than 10 million in SNPs in human genome, and these SNPs are typically
highly conserved within population making SNPs excellent genotypic markers
for medical research. With current microarray technology (SNP-arrays) it is
possible to measure up to one million SNPs in any human tissue sample. Our
objectives are to develop computational tools to efficiently analyze
SNP-array data in cancers and to identify combinatorially acting genes. The
underlying principle of our efforts is to maximize sensitivity (i.e.
maximize the probability that disease associated variants, if any, are
found) and then integrate data from biological databases and published
studies to exclude false positives.
High-throughput data analysis
In many diseases, genes or proteins playing key roles in disease
progression, drug resistance and decreased survival are poorly known.
High-throughput methods have become standard technology in identifying
genes and proteins that are most likely involved in disease initiation,
progression and drug resistance. We have a strong track-record in
experimental design, preprocessing and analyzing data from a number of
high-throughput technologies such as gene, exon, SNP, copy-number
microarrays, mass spectrometry and Illumina Solexa deep-sequencing
platform. Our objective is to translate the raw high-throughput data into
experimentally testable predictions that increase knowledge on disease
mechanisms and effective means of treatment.
Data integration
Data integration is the key element in translating high-throughput data
into trustworthy predictions. Integration is hindered by two challenges:
(i) data come from different sources in various (incompatible) formats and
(ii) software for bioinformatics often does not work as promised and/or is
challenging to use. To overcome these challenges we have developed an open
source component-based workflow framework called
Anduril
for data analysis and integration. The components can be implemented with any
language, such as Java, Python, MATLAB, C++, Weka and R. Anduril automatically produces
reports consisting of all methods, their parameters, figures, tables and so on.
We currently provide components for expression, SNP,
ChIP-chip and exon microarray analysis, and third parties can develop new
components. Implemented database connections include Ensembl, KEGG,
PINA and others.
Signaling pathway modeling
The vast majority of diseases are appreciated to be "complex"; They
arise from alterations within multiple molecular regulatory pathways.
Signaling pathways represent a crucial domain for
pathological dysregulation as they contain forward and reverse feedback
cascades that can act as signal amplifiers, transmitters, or distributors
to a multitude of highly-connected protein nodes across numerous pathways
within a network. We develop and use mechanistic (e.g. differential
equation) and data-driven (e.g. decision tree) based modeling approaches
to characterize the impact of genetic variation, gene/protein
expression patterns, and inhibitors on signaling pathways. At the moment, our
focus is on modeling EGFR, NF-kB and Fas pathways in cancer cells.
Image analysis
Due to advances in cell imaging, image processing is getting increasingly
important in systems biology. Our efforts focus on efficient and automated
analysis of fluorescence imaging, live cell imaging and using the
information extracted from images in statistical analysis. We belong to the
FLUODIAMON
consortium funded by EU FP7.
The overall objective of FLUODIAMON is to develop and validate a quantitative, minimally invasive
diagnostic tool for early and conclusive detection, diagnosis and
monitoring of disease and disease progression of breast and prostate
cancer, with negligible sampling-related side-effects. We are responsible
for bioinformatics analysis in the consortium.
|
