Transcript Bioinformatics - University of Maine System

Bioinformatics
GIS Applications
Anatoly Petrov
Bioinformatics

(in a strict sense) a branch of science dealing with storage, retrieval and
analysis or prediction of the composition or the structure of biomolecules
(sequence analysis)
- nucleic acids (DNA, RNA) - genomics
- proteins - proteomics

(in a wider sense) the intersection of biology and computer science (eg.,
computational biogeography)
Bioinformatics Institute (VBI) sponsored a
major conference focused on the interface
between GIS and bioinformatics: “GIS
Applications to Bioinformatics” (May 16–17,
2001, Blacksburg, Virginia).
A Gene Map of the Human Genome
The Human Transcript Map
Chromosome X
3-dimentional reconstruction
of the amicyanin -
an enzyme participating
in respiration
Chromosome structure
Nucleosome
Chromosomal DNA is packaged into a compact structure with the help of specialized
proteins called histones. The fundamental packing unit is known as a nucleosome.
Sequence features that appear to be spatially disconnected according to a linear
representation of a genome, may actually be close neighbors due to the folding of DNA
into a 3-dimensional molecule.
GenoSIS - Genome Spatial Information
System
(ESRI ArcGIS – visualization tool
+
Oracle Spatial – object-relational database)
Applications:
-
thematic mapping and visualization
exploratory spatial data analysis
(ESDA)
Set of questions for the ESDA
·
Where do we find consensus sequence elements (CSEs)? How many elements are
there at that genomic region?
·
Is there regularity in their distribution? What is the nature of that regularity? Why
should the spatial distributional pattern exhibit regularity?
·
Are CSEs found throughout the genome? What are the limits to where they are
found? Why do those limits constrain its distribution?
·
Are there regulatory elements spatially associated with a gene with a particular
molecular function? Do these regulatory elements and genes usually occur
together in the same places? Why should they be spatially associated?
·
Has a particular gene always been there? In which organism did it first emerge or
become obvious? How has it changed spatially (through evolutionary time)?
·
What factors have influenced its duplication or deletion in the genome? What
factors have constrained its spread?
Using GIS for thematic genome mapping
Application of ArcView for genome mapping
and spatial analysis
Biogeography
Prediction (reconstruction) of species distribution
- In the past (paleobiogeography and paleoclimatology;
ex., NOAA’s Paleoclimatology Program)
- At present (eg., Environment Australia’s Species Mapper)
- In future (eg., some of the Lifemapper products)
Methods
There are two main datasets that are fundamental in obtaining good prediction on a
species distribution map: species occurrence data and environmental
information.
Algorithms
GARP, environmental
classification method)
envelope
(BIOCLIM),
e-ball,
“image”
(Bayesian
Habitat Digitizer Extension
(HDE) to ArcView
uses a hierarchical classification scheme to delineate
habitats by visually interpreting georeferenced images
such as aerial photographs, satellite images, and side
scan sonar.
HDE allows users to create custom classification schemes and rapidly
delineate and attribute polygons using simple menus.
Deducing potential species distribution
using BIOCLIM
Environment Australia’s Species Mapper
Query database to retrieve records of
species locations.
 For each species location, interpolate
values of essential climatic variables.

Calculate the climatic envelope bounding
all the species records.
 At the resolution specified, identify all
other sites in the landscape that fall
within the climatic envelope.


Plot the sites identified on a base map.

Deliver the map to the user.
Lifemapper
1. The species occurrence data is gathered from a number of
biological collections housed at several museums and herbaria
worldwide. Those institutions have their specimen databases
linked and integrated through The Species Analyst project.
2. The environmental information is represented as a set of
geographic layers. Each layer displays one particular
environmental parameter, such as temperature, rainfall, land use,
elevation, among others.
3. Using data from those two datasets, GARP tries to find
nonrandom correlations between species occurrence data and
the values of the environmental parameters where the species
occur or do not occur.
G Genetic
A Algorithm for
R Ruleset
P Production
Paleobiogeography
3-D Flythrough
animation
Holocene Evolution
of the Southern
Washington
and
Northern
Oregon
Shelf and Coast.
NOAA’s Paleoclimatology Program
Pollen Viewer
THE END