Clustering and classification in gene expression data
1 / 55

Clustering and Classification In Gene Expression Data - PowerPoint PPT Presentation

  • Updated On :

Clustering and Classification In Gene Expression Data. Carlo Colantuoni [email protected] Slide Acknowledgements:

Related searches for Clustering and Classification In Gene Expression Data

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Clustering and Classification In Gene Expression Data' - irma

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Clustering and classification in gene expression data l.jpg

Clustering and Classification In Gene Expression Data

Carlo Colantuoni

[email protected]

Slide Acknowledgements:

Elizabeth Garrett-Mayer, Rafael Irizarry, Giovanni Parmigiani, David Madigan, Kevin Coombs, Richard Simon, Ingo Ruczinski.Classification based in part on Chapter 10 of Hand, Manilla, & Smyth and Chapter 7 of Han and Kamber

Slide2 l.jpg

Data from Garber et al.

PNAS (98), 2001.

Slide4 l.jpg


  • Clustering is an exploratory tool to see who's running with who: Genes and Samples.

  • “Unsupervized”

  • NOT for classification of samples.

  • NOT for identification of differentially expressed genes.

Clustering l.jpg

  • Clustering organizes things that are close into groups.

  • What does it mean for two genes to be close?

  • What does it mean for two samples to be close?

  • Once we know this, how do we define groups?

  • Hierarchical and K-Means Clustering

Distance l.jpg

  • We need a mathematical definition of distance between two points

  • What are points?

  • If each gene is a point, what is the mathematical definition of a point?

Points l.jpg

1 2 . . . . . . . N












  • Gene1= (E11, E12, …, E1N)’

  • Gene2= (E21, E22, …, E2N)’

  • Sample1= (E11, E21, …, EG1)’

  • Sample2= (E12, E22, …, EG2)’

  • Egi=expression gene g, sample i


Most famous distance l.jpg
Most Famous Distance

  • Euclidean distance

    • Example distance between gene 1 and 2:

    • Sqrt of Sum of (E1i -E2i)2, i=1,…,N

  • When N is 2, this is distance as we know it:




When N is 20,000 you have to think abstractly

Correlation can also be used to compute distance l.jpg
Correlation can also be used to compute distance

  • Pearson Correlation

  • Spearman Correlation

  • Uncentered Correlation

  • Absolute Value of Correlation

Slide10 l.jpg

The difference is that, if you have two vectors X and Y with identical

shape, but which are offset relative to each other by a fixed value,

they will have a standard Pearson correlation (centered correlation)

of 1 but will not have an uncentered correlation of 1.

The similarity distance matrices l.jpg
The similarity/distance matrices identical

1 2 ………………………………...G

1 2 ……….N

























The similarity distance matrices12 l.jpg
The similarity/distance matrices identical

1 2 ……….N

1 2 …………..N




















Gene and sample selection l.jpg
Gene and Sample Selection identical

  • Do you want all genes included?

  • What to do about replicates from the same individual/tumor?

  • Genes that contribute noise will affect your results.

  • Including all genes: dendrogram can’t all be seen at the same time.

  • Perhaps screen the genes?

Two commonly seen clustering approaches in gene expression data analysis l.jpg
Two commonly seen clustering approaches in gene expression data analysis

  • Hierarchical clustering

    • Dendrogram (red-green picture)

    • Allows us to cluster both genes and samples in one picture and see whole dataset “organized”

  • K-means/K-medoids

    • Partitioning method

    • Requires user to define K = # of clusters a priori

    • No picture to (over)interpret

Hierarchical clustering l.jpg
Hierarchical Clustering data analysis

  • The most overused statistical method in gene expression analysis

  • Gives us pretty red-green picture with patterns

  • But, pretty picture tends to be pretty unstable.

  • Many different ways to perform hierarchical clustering

  • Tend to be sensitive to small changes in the data

  • Provided with clusters of every size: where to “cut” the dendrogram is user-determined

Choose clustering direction l.jpg
Choose clustering direction data analysis

  • Agglomerative clustering (bottom-up)

    • Starts with as each gene in its own cluster

    • Joins the two most similar clusters

    • Then, joins next two most similar clusters

    • Continues until all genes are in one cluster

  • Divisive clustering (top-down)

    • Starts with all genes in one cluster

    • Choose split so that genes in the two clusters are most similar (maximize “distance” between clusters)

    • Find next split in same manner

    • Continue until all genes are in single gene clusters

Choose linkage method if bottom up l.jpg
Choose linkage method (if bottom-up) data analysis

  • Single Linkage: join clusters whose distance between closest genes is smallest (elliptical)

  • Complete Linkage: join clusters whose distance between furthest genes is smallest (spherical)

  • Average Linkage: join clusters whose average distance is the smallest.

Slide21 l.jpg

Cluster Assignment data analysis

Slide22 l.jpg

Simulated Data with 4 clusters: data analysis1-10, 11-20, 21-30, 31-40

450 relevant genes + 450 “noise” genes.

450 relevant genes.

K means and k medoids l.jpg
K-means and K-medoids data analysis

  • Partitioning Method

  • Don’t get pretty picture

  • MUST choose number of clusters K a priori

  • More of a “black box” because output is most commonly looked at purely as assignments

  • Each object (gene or sample) gets assigned to a cluster

  • Begin with initial partition

  • Iterate so that objects within clusters are most similar

K means continued l.jpg
K-means (continued) data analysis

  • Euclidean distance most often used

  • Spherical clusters.

  • Can be hard to choose or figure out K.

  • Not unique solution: clustering can depend on initial partition

  • No pretty figure to (over)interpret

K means algorithm l.jpg
K-means Algorithm data analysis

1. Choose K centroids at random

2. Make initial partition of objects into k clusters by assigning objects to closest centroid

  • Calculate the centroid (mean) of each of the k clusters.

  • a. For object i, calculate its distance to each of

    the centroids.

    b. Allocate object i to cluster with closest


    c. If object was reallocated, recalculate centroids based

    on new clusters.

    4. Repeat 3 for object i = 1,….N.

  • Repeat 3 and 4 until no reallocations occur.

  • Assess cluster structure for fit and stability

K means l.jpg

We start with some data data analysis


We are showing expression for two samples for 14 genes

We are showing expression for two genes for 14 samples

This is with 2 genes.


Iteration = 0

K means27 l.jpg

Choose K data analysiscentroids

These are starting values that the user picks.

There are some data driven ways to do it


Iteration = 0

K means28 l.jpg

Make first data analysispartition by finding the closest centroid for each point

This is where distance is used


Iteration = 1

K means29 l.jpg

Now re-compute the centroids by taking the data analysismiddle of each cluster


Iteration = 2

K means30 l.jpg

Repeat until the centroids stop moving or until you get tired of waiting


Iteration = 3

K means limitations l.jpg
K-means Limitations tired of waiting

  • Final results depend on starting values

  • How do we chose K? There are methods but not much theory saying what is best.

  • Where are the pretty pictures?

Assessing cluster fit and stability l.jpg
Assessing cluster fit and stability tired of waiting

  • Most often ignored.

  • Cluster structure is treated as reliable and precise

  • Can be VERY sensitive to noise and to outliers

  • Homogeneity and Separation

  • Cluster Silhouettes: how similar genes within a cluster are to genes in other clusters (Rousseeuw Journal of Computation and Applied Mathematics, 1987)

Silhouettes l.jpg
Silhouettes tired of waiting

  • Silhouette of gene i is defined as:

  • ai = average distance of gene i to other gene in same cluster

  • bi = average distance of gene i to genes in its nearest neighbor cluster

Wadp weighted average discrepancy pairs l.jpg

Add perturbations to original data tired of waiting

Calculate the number of paired samples that cluster together in the original cluster that didn’t in the perturbed

Repeat for every cutoff (i.e. for each k)

Do iteratively

Estimate for each k the proportion of discrepant pairs.

WADP: Weighted Average Discrepancy Pairs

Slide36 l.jpg

Classification tired of waiting

  • Diagnostic tests are good examples of classifiers

    • A patient has a given disease or not

    • The classifier is a machine that accepts some clinical parameters as input, and spits out an prediction for the patient

      • D

      • Not-D

  • Classes must be mutually exclusive and exhaustive

Components of class prediction l.jpg
Components of Class Prediction tired of waiting

  • Select features (genes)

    • Which genes will be included in the model

  • Select type of classifier

    • E.g. (D)LDA, SVM, k-Nearest-Neighbor, …

  • Fit parameters for model (train the classifier)

  • Quantify predictive accuracy: Cross-Validation

Feature selection l.jpg
Feature Selection tired of waiting

  • Goal is to identify a small subset of genes which together give accurate predictions.

  • Methods will vary depending on nature of classification problem

    • Choose genes with significant t-statistics to distinguish between two simple classes e.g.

Classifier selection l.jpg
Classifier Selection tired of waiting

  • In microarray classification, the number of features is (almost) always much greater than the number of samples.

  • Overfitting is a distinct risk, and increases with more complicated methods.

How microarrays differ from the rest of the world l.jpg
How microarrays differ from the rest of the world tired of waiting

  • Complex classification algorithms such as neural networks that perform better elsewhere don’t do as well as simpler methods for expression data.

  • Comparative studies have shown that simpler methods work as well or better for microarray problems because the number of candidate predictors exceeds the number of samples by orders of magnitude.

    (Dudoit, Fridlyand and Speed JASA 2001)

Statistical methods appropriate for class comparison may not be appropriate for class prediction l.jpg
Statistical Methods Appropriate for Class Comparison may not be Appropriate for Class Prediction

  • Demonstrating statistical significance of prognostic factors is not the same as demonstrating predictive accuracy.

  • Demonstrating goodness of fit of a model to the data used to develop it is not a demonstration of predictive accuracy.

  • Most statistical methods were not developed for p>>n prediction problems

Linear discriminant analysis l.jpg
Linear discriminant analysis be Appropriate for Class Prediction

  • If there are K classes, simply draw lines (planes) to divide the space of expression profiles into K regions, one for each class.

  • If profile X falls in region K, predict class K.

Nearest neighbor classification l.jpg
Nearest Neighbor Classification be Appropriate for Class Prediction

  • To classify a new observation X, measure the distance d(X,Xi) between X and every sample Xi in training set

  • Assign to X the class label of its “nearest neighbor” in the training set.

Random forests l.jpg
Random Forests be Appropriate for Class Prediction

Build several “random” decision trees and have them vote to determine final classification

Evaluating a classifier l.jpg
Evaluating a classifier be Appropriate for Class Prediction

  • Want to estimate the error rate when classifier is used to predict class of a new observation

  • The ideal approach is to get a set of new observations, with known class label and see how frequently the classifier makes the correct prediction.

  • Performance on the training set is a poor approach, and will deflate the error estimate.

  • Cross validation methods are used to get less biased estimates of error using only the training data.

Split sample evaluation l.jpg
Split-Sample Evaluation be Appropriate for Class Prediction

  • Training-set

    • Used to select features, select model type, determine parameters and cut-off thresholds

  • Test-set

    • Withheld until a single model is fully specified using the training-set.

    • Fully specified model is applied to the expression profiles in the test-set to predict class labels.

    • Number of errors is counted

V fold cross validation l.jpg
V-fold cross validation be Appropriate for Class Prediction

  • Divide data into V groups.

  • Hold one group back, train the classifier on other V-1 groups, and use it to predict the last one.

  • Rotate through all V points, holding each back.

  • Error estimate is total error rate on all V test groups.

Leave one out cross validation l.jpg
Leave-one-out Cross Validation be Appropriate for Class Prediction

  • Hold one data point back, train the classifier on other n-1 data points, and use it to predict the last one.

  • Rotate through all n points, holding each back.

  • Error estimate is total error rate on all n test values.

Slide49 l.jpg

log-expression ratios be Appropriate for Class Prediction

full data set


log-expression ratios

training set


test set

Non-cross-validated Prediction

1. Prediction rule is built using full data set.

2. Rule is applied to each specimen for class prediction.

Cross-validated Prediction (Leave-one-out method)

1. Full data set is divided into training and test sets (test set contains 1 specimen).

2. Prediction rule is built from scratch using the training set.

3. Rule is applied to the specimen in the test set for class prediction.

4. Process is repeated until each specimen has appeared once in the test set.

Which to use depends mostly on sample size l.jpg
Which to use depends mostly on sample size be Appropriate for Class Prediction

  • If the sample is large enough, split into test and train groups.

  • If sample is barely adequate for either testing or training, use leave one out

  • In between consider V-fold. This method can give more accurate estimates than leave one out, but reduces the size of training set.

Beware l.jpg
Beware be Appropriate for Class Prediction

  • Cross-validation of a model cannot occur after selecting the genes to be used in the model

Incomplete incorrect cross validation l.jpg
Incomplete (incorrect) Cross-Validation be Appropriate for Class Prediction

  • Publications are using all the data to select genes and then cross-validating only the parameter estimation component of model development

    • Highly biased

    • Many published complex methods which make strong claims based on incorrect cross-validation.

      • Frequently seen in complex feature set selection algorithms

      • Some software encourages inappropriate cross-validation

Slide53 l.jpg

cDNA Microarrays be Appropriate for Class Prediction

Parallel Gene Expression Analysis

Gene-Expression Profiles in Hereditary Breast Cancer

  • Breast tumors studied:

    • 7 BRCA1+ tumors

    • 8 BRCA2+ tumors

    • 7 sporadic tumors

  • Log-ratios measurements of 3226 genes for each tumor after initial data filtering


Can we distinguish BRCA1+ from BRCA1– cancers and BRCA2+ from BRCA2– cancers based solely on their gene expression profiles?

Brca1 l.jpg
BRCA1 be Appropriate for Class Prediction

Brca2 l.jpg
BRCA2 be Appropriate for Class Prediction