Title: Computational Gene Finding
1Computational Gene Finding
CIS786 Intro to Comp Biol Instructor Dr. Barry
Cohen
- Greg Voronin
- Hui Zhao
- Xueyi(Judy) Xiao
2The Challenge
Presented By Greg Voronin
- Generate predictions of gene locations from
primary genomic sequence by computational means - Two principle means
- Database searching
- Statistical Methods
3The Biological Model
4The Computational Model
- Representing the biology in a framework amenable
to mathematical/statistical methods - Exon classification, sequence features, signal
profiles - What is an exon and what properties does the
sequence of an exon hold? - How is an exon recognized and processed?
5Exon Classification Scheme
6The Nature of The Data
- What is the primary genomic sequence?
- Nor is the available sequence a single
continuous and exact sequence for each
chromosome the HGP is represented by a set
of sequences that cover the genome is a
statistical sense but have a very large number of
gaps. - Many genes are as large or larger than the
contigs in the HGP - Finding genes will depend on the accuracy of the
scaffold of their contigs
7Back to Beginning
- What is a gene?
- A biological model, a mathematical model and
computational representation
The programs we evaluate take these factors into
account in their underlying model
8MZEF
- Michael Zhangs Exon Finder
- Utilizes quadratic discriminant analysis (QDA) to
classify sequence into gene and non-gene groups - QDA is a multivariate statistical pattern
recognition method - Draws a curved boundary between groups of
different classes
9QDA
10Key Elements of QDA
- Entities are represented by an n-dimensional
vector of feature values - Two classes of entities are categorized by their
respective multinormal distribution - Each class has its own mean vector
- The mean of each feature
- An appropriate distance function is central to
the calculation of the posterior probabillity of
group membership of a given unknown entity given
its specific feature vector.
11Mahalanobis Distance
- The actual posterior probabillity function is
more complex, but this is the distance component
( x mi )T Si-1 ( x mi )
12MZEF Specifics
- MZEF uses the following features
- Exon length, exon-intron transition, branch site
score, 3ss score, exon score, strand score,
frame score, 5ss score, intron-exon transition - 9 dimensional feature vector
- Training sets of known exons and non-exons are
used to establish the class characterisitics - Supervised learning
13GATC to Gene
- Cells recognize genes from DNA sequence.
Can we??
The Hidden Markov Model Method
HMMgene Presented By Hui Zhao
14HMMs are Statistical Models
- Definition
- Any mathematical construct that attempts to
parameterize a random process - Example A normal distribution
- Assumptions
- Parameters
- Estimation
- Usage
- HMMs are just a little more complicated
15Primary HMM Assumptions
- Observations are ordered
- Random processes can be represented by a
stochastic finite state machine with emitting
states - transition probabilities and emission
probabilities.
16How do we find the model probabilities?
- This is called training
- We start with an architecture and a set of
observed sequences - The training process iteratively alters its
parameters to fit the training set - The trained model will assign the training
sequences high probability - but can it generalize?
17HMM Usage two major tasks
- Evaluate the probability of an observed sequence
given the model (Forward) - Find the most likely path through the model for a
given observation sequence (Viterbi)
18Gene Finding An Ideal HMM Application
- Our Objective
- To find the coding and non-coding regions of an
unlabeled string of DNA nucleotides - Our Motivation
- Assist in the annotation of genomic data produced
by genome sequencing methods - Gain insight into the mechanisms involved in
transcription, splicing and other processes
19Why HMMs might be a good fit for Gene Finding
- The observations within a sequence are ordered
- A DNA sequence is a set of ordered observations
- Designing the architecture is straight forward
- Easy to measure success
- Training data is available from various genome
annotation projects
20A HMM genefinder
- States represent standard gene features
intergenic region, exon, intron, perhaps more
(promotor, 5UTR, 3UTR, Poly-A,..). - Observations are things like state-dependent
base composition. - In a HMM, length of each state must be included
as well. -
- Finally, reading frames and both strands must be
dealt with.
21Several problems can occur
5
correct gene structure extended exon missing
exon additional exon missing intron extended gene
model
22HMMgene
Krogh (1997) In Proc. 5th Conf. Intel. Sys. Mol
Biol. pp179-186
23HMMGene
- Uses an extended HMM called a CHMM
- CHMM HMM with classes
- Takes full advantage of being able to modify the
statistical algorithms - Uses high-order states
- Trains everything at once
24How does HMMGene work?
1) 5th order HMM assumes P(xi xi-1,xi-2,
xi-3, xi-4, xi-5) is different in Introns, Exons,
etc..
2) Construct the model
252. How does HMMGene work?
4) Use Viterbi (n-best) to find a path
through the CHMM a labeled gene
5) Use the forward algorithm to measure P(gene
model) using n-best.
26A DNA sequence containing one gene. For each
nucleotide its label is written below. The coding
regions are labeled C, the introns I, and the
intergenic regions 0. HMMGene calls these
class labels in a CHMM.
27HMMGene
- Does not use the standard ML method which
optimizes the probability of the observed
sequence instead it maximizes the probability
of the correct prediction. - Only one conference paper describes the
algorithm. There is a web site to run the
algorithm, and it's performance has been compared
to other algorithms. - No complete description of the algorithm is
available in the 1997 paper the author states
" the details of HMMGene will be described
elsewhere (in prep)" but unfortunately the
detailed paper has not been published.
28HMMgene http//www.cbs.dtu.dk/services/HMMgene/)
29HMMgene and HMM Disadvantages
- Markov Chains
- States should be independent
- P(y) must be independent of P(x) -usually not
true - Local maxima
- Model may not converge the optimal parameter set
- Over-fitting
- More training is not always good-set may be too
small
30Summary
- HMMgene finds whole genes in anonymous DNA with
correctly spliced exons. - It can predict several whole or partial genes in
one sequence. - If some features of a sequence are known, such as
hits to ESTs, proteins, or repeat elements, these
regions can be locked as coding or non-coding and
then the program will find the best gene
structure under these constraints.
31GENSCAN (v1.0)
Presented By Xueyi (Judy) Xiao
- A computer program identifying complete exon
intron structures of genes in genomic DNA. - Developed by Chris Burge (Burge 1997), in the
research group of Samuel Karlin, Dept of
Mathematics, Stanford Univ. - Original server _at_Stanford ? New server _at_MIT
(seq_len lt 500 kb) - Servers are also maintained by the Pasteur
Institute, Paris and by the GENSCAN web server at
DKFZ/EMBnet, Heidelberg - Implementations
- web server http//genes.mit.edu/GENSCAN.html
- email server http//genes.mit.edu/GENSCANM.html
- local copy downloaded under a license agreement
32How does It Work?
- Designed to predict complete gene structures
- Introns and exons
- Promoter sites
- Polyadenylation signals
- Larger predictive scope
- Partial and Complete genes
- Multiple genes separated by intergenic DNA in a
seq - Consistent sets of genes on either/both DNA
strands - Not use similarity-based methods
- Based on a general probabilistic model of genomic
sequences composition and gene structure
33Model of Genomic Sequence Structure
Fig. 3, Burge and Karlin 1997
34Input
http//genes.mit.edu/GENSCAN.html
35Output
36Graphic View
Optimal Exon Suboptimal Exon
Initial Exon
Internal Exon
Terminal Exon
Single-Exon gene
37Is It Good?
- Accuracy
- Substantially higher accuracies when tested on
standardized sets of human vertebrate genes,
with 75-80 of exons identified exactly. - Reliability
- Able to indicate fairly accurately the
reliability of each predicted exon. - Consistency
- Consistently high levels of accuracy, for seqs
of differing CG content and for distinct groups
of vertebrates.
38Why not Perfect?
- Gene Number
- usually approximately correct, but may not
- Organism
- primarily for human/vertebrate seqs maybe lower
accuracy for non-vertebrates. Glimmer
GeneMark for prokaryotic or yeast seqs -
- Exon and Feature Type
- Internal exons gt Initial or Terminal exons
- Exons gt Polyadenylation or Promoter
signals(NNPP) -
- Biases in Test Set
- The Burset/Guigó (1996) dataset
- toward short genes with relatively simple
exon/intron structure - The Rogic (2001) dataset
- DNA seqs GenBank r-111.0 (04/1999 lt- 08/1997)
- source organism specified
- consider genomic seqs containing exactly one
gene - seqsgt200kb were discarded mRNA seqs and seqs
containing pseudo genes or alternatively spliced
genes were excluded.
39What are They doing NOW?
- The research group _at_MIT
- is currently developing another program,
GenomeScan, which is more accurate - when a moderate or closely related
- protein seq is available.
40(No Transcript)
41TEST OF METHODS
- Sample Tests reported by Literature
- Test on the set of 570 vertebrate gene seqs
(BursetGuigo 1996) as a standard for comparison
of gene finding methods. - Test on the set of 195 seqs of human, mouse or
rat origin (named HMR195) (Rogic 2001). - Self-Test done by our group
- Dataset Intron-less(Single-exon),
-rich(Multi-exon), -poor(Random) - Organism Human
- Methods all of the three
- Steps
42Where to get the dataset for Self-Test?
http//www.ncbi.nlm.nih.gov/genome/guide/human/
43Accuracy Measures
Sensitivity vs. Specificity (adapted from
BursetGuigo 1996)
44Results Accuracy Statistics
- Table Relative Performance (adapted added from
Rogic 2001)
of seqs - number of seqs effectively analyzed
by each program in parentheses is the number of
seqs where the absence of gene was predicted Sn
-nucleotide level sensitivity Sp - nucleotide
level specificity CC - correlation coefficient
ESn - exon level sensitivity ESp - exon level
specificity
45Testing Random Sequences
Presented By Greg Voronin
- These gene finding programs model statistical
trends and properties - Can they be fooled by random sequences
- Generate a preliminary measure of accuracy
- Java program written to generate random
sequences of a,t,g,c - 3 groups of sequences 5k, 10k 30K
- Sent to BLAST then GeneMachine
46Testing Results
- BLAST
- bit score E-value
- 5k 42 5.7
- 10k 44 3.0
- 30k 42 8.7
- GeneMachine
- 5k 10k 30K
- MZEF 1 5 14
- GenScan 3 11 26
- HMMgene 7 11 42
47New directions
Presented By Hui Zhao
- Computational Gene Finding has rapidly evolved
since it started 20 years ago. - The advent of full-length genomic sequences has
provided data and increased the requirements. - Gene annotation has direct medical implications
on the design of pharmaceuticals and the
understanding of the genetic component of
diseases. - Gene finding remains largely an unsolved problem.
48New directions
- The growing quantities of training data for the
models should improve their performance. - Algorithms that combine the inputs from several
models in a weighted voting scheme should be
considered to try to get the best from all of the
methods. - Many other AI approaches can be used to meet this
challenge including decision trees, neural
networks and rule-based systems
49Challenges and Discoveries Ahead
- Eukaryotic gene finding continues to be an active
and important area more research is required
into algorithms with greater accuracy - Expertise in computational biology is also
required which means training in both computer
science and molecular biology - More classes like this