DNA Identification: Bayesian Belief Update Cybergenetics © 2003-2010 TrueAllele ® Lectures Fall, 2010 Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh,

Slides:

Advertisements

Similar presentations

DNA Identification: Mixture Weight & Inference

Advertisements

Overcoming DNA Stochastic Effects 2010 NEAFS & NEDIAI Meeting November, 2010 Manchester, VT Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh, PA Cybergenetics.

Sherlock Holmes and the DNA Likelihood Ratio American Academy of Forensic Sciences February, 2011 Chicago, IL Mark W Perlin, PhD, MD, PhD Cybergenetics,

Forensic DNA Inference ICFIS 2008 Lausanne, Switzerland Mark W Perlin, PhD, MD, PhD Joseph B Kadane, PhD Robin W Cotton, PhD Cybergenetics ©

How Inclusion Interpretation of DNA Mixture Evidence Reduces Identification Information American Academy of Forensic Sciences February, 2013 Washington,

How TrueAllele ® Works (Part 3) Kinship, Paternity and Missing Persons Cybergenetics Webinar December, 2014 Mark W Perlin, PhD, MD, PhD Cybergenetics,

Bayesian inference “Very much lies in the posterior distribution” Bayesian definition of sufficiency: A statistic T (x 1, …, x n ) is sufficient for 

Psychology 290 Special Topics Study Course: Advanced Meta-analysis April 7, 2014.

Binomial Distribution & Bayes’ Theorem. Questions What is a probability? What is the probability of obtaining 2 heads in 4 coin tosses? What is the probability.

Intro to Bayesian Learning Exercise Solutions Ata Kaban The University of Birmingham 2005.

Basics of Statistical Estimation. Learning Probabilities: Classical Approach Simplest case: Flipping a thumbtack tails heads True probability  is unknown.

CSE 221: Probabilistic Analysis of Computer Systems Topics covered: Statistical inference (Sec. )

CSE 221: Probabilistic Analysis of Computer Systems Topics covered: Statistical inference.

CENTER FOR BIOLOGICAL SEQUENCE ANALYSIS Bayesian Inference Anders Gorm Pedersen Molecular Evolution Group Center for Biological Sequence Analysis Technical.

Introduction to Bayesian statistics Three approaches to Probability  Axiomatic Probability by definition and properties  Relative Frequency Repeated.

How TrueAllele® Works (Part 1)

Additional Slides on Bayesian Statistics for STA 101 Prof. Jerry Reiter Fall 2008.

A quick intro to Bayesian thinking 104 Frequentist Approach 10/14 Probability of 1 head next: = X Probability of 2 heads next: = 0.51.

Lecture 5a: Bayes’ Rule Class web site: DEA in Bioinformatics: Statistics Module Box 1Box 2Box 3.

Naive Bayes Classifier

Bayesian Hypothesis Testing for Proportions Antonio Nieto / Sonia Extremera / Javier Gómez PhUSE Annual Conference, 9th-12th Oct 2011, Brighton UK.

CSE 446: Point Estimation Winter 2012 Dan Weld Slides adapted from Carlos Guestrin (& Luke Zettlemoyer)

Week 71 Hypothesis Testing Suppose that we want to assess the evidence in the observed data, concerning the hypothesis. There are two approaches to assessing.

Bayes Theorem Thomas R. Stewart, Ph.D. Center for Policy Research Rockefeller College of Public Affairs and Policy University at Albany State University.

Likelihood function and Bayes Theorem In simplest case P(B|A) = P(A|B) P(B)/P(A) and we consider the likelihood function in which we view the conditional.

Bayesian vs. frequentist inference frequentist: 1) Deductive hypothesis testing of Popper--ruling out alternative explanations Falsification: can prove.

Bayesian Networks for Cyber Crimes. Bayes’ Theorem For an hypothesis H supported by evidence E: Pr(H|E) = Pr(E|H).Pr(H)/Pr(E) where – Pr(H|E) is the posterior.

How TrueAllele ® Works (Part 2) Degraded DNA and Allele Dropout Cybergenetics Webinar November, 2014 Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh,

Cybergenetics Webinar January, 2015 Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh, PA Cybergenetics © How TrueAllele ® Works (Part 4)

Prediction statistics Prediction generally True and false, positives and negatives Quality of a prediction Usefulness of a prediction Prediction goes Bayesian.

What’s in the Bag? Lecture 4 Section Wed, Jan 25, 2006.

Bayesian Statistics and Decision Analysis

Getting Past First Bayes with DNA Mixtures American Academy of Forensic Sciences February, 2014 Seattle, WA Mark W Perlin, PhD, MD, PhD Cybergenetics,

Bayesian Prior and Posterior Study Guide for ES205 Yu-Chi Ho Jonathan T. Lee Nov. 24, 2000.

Murder in McKeesport October 25, 2008 Tamir Thomas.

MPS/MSc in StatisticsAdaptive & Bayesian - Lect 71 Lecture 7 Bayesian methods: a refresher 7.1 Principles of the Bayesian approach 7.2 The beta distribution.

Open Access DNA Database Duquesne University March, 2013 Pittsburgh, PA Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh, PA Cybergenetics ©

The Uniform Prior and the Laplace Correction Supplemental Material not on exam.

Bayes Theorem, a.k.a. Bayes Rule

Objective DNA Mixture Information in the Courtroom: Relevance, Reliability & Acceptance NIST International Symposium on Forensic Science Error Management:

Statistical NLP: Lecture 4 Mathematical Foundations I: Probability Theory (Ch2)

DNA Identification: Quantitative Data Modeling Cybergenetics © Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh, PA TrueAllele ® Lectures.

Outline Historical note about Bayes’ rule Bayesian updating for probability density functions –Salary offer estimate Coin trials example Reading material:

Naive Bayes Classifier. REVIEW: Bayesian Methods Our focus this lecture: – Learning and classification methods based on probability theory. Bayes theorem.

Introduction to Bayesian inference POORFISH workshop Helsinki Samu Mäntyniemi Fisheries and Environmental Management group (FEM) Department.

MSc Methods part II: Bayesian analysis Dr. Mathias (Mat) Disney UCL Geography Office: 113, Pearson Building Tel:

Bayesian Estimation and Confidence Intervals Lecture XXII.

Intro to Bayesian Learning Exercise Solutions Ata Kaban The University of Birmingham.

Bayesian Estimation and Confidence Intervals

A Match Likelihood Ratio for DNA Comparison

Validating TrueAllele® genotyping on ten contributor DNA mixtures

Naive Bayes Classifier

Error in the likelihood ratio: false match probability

Bayes Net Learning: Bayesian Approaches

Explaining the Likelihood Ratio in DNA Mixture Interpretation

Null Hypothesis Testing

Binomial Distribution & Bayes’ Theorem

An Early Detection and Rapid Assessment Network for Plant Invasions

Solving Crimes using MCMC to Analyze Previously Unusable DNA Evidence

DNA Identification: Inclusion Genotype and LR

Statistical NLP: Lecture 4

DNA Identification: Stochastic Effects

DNA Identification: Biology and Information

DNA Identification: Biology and Information

Forensic match information: exact calculation and applications

Lecture 4 Section Wed, Sep 6, 2006

Wellcome Trust Centre for Neuroimaging

Bayes for Beginners Luca Chech and Jolanda Malamud

DNA Identification: Biology and Information

DNA Identification: Mixture Interpretation

Presentation transcript:

DNA Identification: Bayesian Belief Update Cybergenetics © TrueAllele ® Lectures Fall, 2010 Mark W Perlin, PhD, MD, PhD Cybergenetics, Pittsburgh, PA

Prior Probability What do we believe before we see any data?

Likelihood Function How well does each hypothesis explain the data?

Posterior Probability What do we believe after we see the data?

Bayes Theorem Our belief in a hypothesis after seeing data is proportional to how well that hypothesis explains the data times our initial belief. All hypotheses must be considered. Need computers to do this properly. Find the probability of causes by examining effects. Rev Bayes, 1763 Computers, 1992

Bayesian Update Posterior probabilityLikelihood functionPrior probability

Bayesian Update Posterior probabilityLikelihood functionPrior probability Consider all possibilities

Parameter Update Beta distribution Binomial distribution Bayes original example

Beta distribution: uniform prior

Weak prior probability of 1/3

Strong prior around 1/4

Binomial likelihood: 1 in 2

Observe: 7 counts in 10 trials

65 counts in 100 trials

Initial belief around 1/3

Observe 1 count in 2 trials

Final belief: posterior probability

Initial belief (prior probability)

Data: 7 counts in 10 trials

Final belief (posterior probability)

Weak prior probability (1/3)

Strong likelihood (65 in 100)

Strong posterior probability

Disease Prevalence With a positive medical test result, what is the chance of having this rare disease? Free of Disease Got the Disease Pr(Free) 99.9% (= 100% - 0.1%) Pr(Got) 0.1% 1 in a thousand

Medical Test: Likelihood Data Free of Disease Got the Disease Positive Test False positive Pr(Pos | Free) 5% 5 in a hundred True positive Pr(Pos | Got) 99% (= 100% - 1%) Negative Test True negative Pr(Neg | Free) 95% (= 100% - 5%) False negative Pr(Neg | Got) 1% 1 in a hundred

Probability of Disease With a positive test (Pos), what is the probability of having the disease (Got)?

Prior: 99.9% disease free

Positive test: 99% true, 5% false

Posterior: 98% disease free

Odds of Disease With a positive test (Pos), what are the odds of having the disease (Got vs. Free)?

Likelihood Ratio Data Free of Disease Got the Disease Positive Test False positive Pr(Pos | Free) 5% 5 in a hundred True positive Pr(Pos | Got) 99% (= 100% - 1%) Negative Test True negative Pr(Neg | Free) 95% (= 100% - 5%) False negative Pr(Neg | Got) 1% 1 in a hundred