Star

Home

Biochem

Biochem

Genetics

Genetics

Orf

Orf

Hydro

Hydro

Molsim

Molsim

Cluster

Cluster

Hpc

Hpc

Instructions for Problem Set

In this problem set you will use three types of viewers to analyze gene expression data. Below are the instructions for logging on to GenePattern and running any one of the viewers (or pipelines). You will be instructed in the problem set below to run a particular pipeline depending on the questions being asked.

View the Original Data

  1. Log on to GenePattern (refer to “Logging on to GenePattern” on page 5 for instructions)
  2. Run the “seeOriginalData” pipeline from GenePattern using the all_aml dataset. (Refer to “How to run a pipeline” on page 5 for instructions)

When you perform a microarray experiment to see what genes are expressed at very high or very low levels in ALL tumors and AML tumors, you get a collection of raw data that looks like this.Original Data Main Window

Original Data Viewer

As you can see, these data are hard to interpret. The program you are about to use allows you to visualize these data in easily interpretable ways. We will examine below three ways in which these data can be visualized.

MicroArray Data

Navigate to the main page by clicking the GenePattern logo at the top left of the page. Run the “seeMicroArrayData” pipeline from GenePattern using the all_aml dataset. (Refer to “How to run a pipeline” on page 5 for instructions)

MicroArray Viewer Main Window

MicroArray Viewer

How to Answer Question e)

You can search for a gene by finding its identification number in the column entitled Feature on the Upregulated Features page, which can be done using the Find command apple-F for Mac or ctrl-F for Windows/Linux:

MicroArray Viewer - Find

How to Answer Questions f) and g)
  • Highlight the identification number of the gene in the spreadsheet.
  • From the menu options, choose File > GeneCruiser
  • Select the following options within the LocusLink group by clicking on the boxes next to them so that the boxes become filled with check marks: Chromosome, Description, and Location
  • After checking these options, click the OK button at the bottom right of the configuration panel.
  • RESULT: In the MicroArray viewer, there are three new columns at the far right entitled: “LocusLink Chromosome”, “LocusLink Description”, and “LocusLink Location”.

HeatMap Data

Navigate to the main page by clicking the GenePattern logo at the top left of the page. Run the “seeHeatMapData” pipeline from GenePattern using the all_aml dataset. (Refer to “How to run a pipeline” on page 5 for instructions).

HeatMap Main Window

HeatMap Main Window

NOTE: To zoom in, go to Edit > Display Options, and increase the grid size by moving the slider bar to the right. To zoom out, move the slider bar to the left.

How to Answer Question h)

The columns, labeled at the top of the heat map shown in the figure above, are the individual tumor samples. Each row, labeled to the right of the heat map, corresponds to the level of expression of a particular gene in all the tumor samples. A red block means a “high” level of expression and blue means a “low” level of expression (similar to hot/cold). Identify which sample behaves least like the others with respect to expression in Terminal Transferase.

Clustering Data

Run the “seeClusterData” pipeline from GenePattern using the all_aml dataset. (Refer to “How to run a pipeline” on page 5 for instructions)

NOTE: To zoom in, go to Edit > Display Options, and increase the grid size by moving the slider bar to the right. To zoom out, move the slider bar to the left.

How to Answer Questions i) and j)

Similar to the heat map viewer, the columns are the individual tumor samples while each row corresponds to a specific gene and shows its level of expression in all the tumor samples. A red block means a “high” level of expression and blue means a “low” level of expression (similar to hot/cold). In this case, the tumor samples (columns) are “clustered” or grouped according to how similar their expressions are. Use this to determine which sample is least like the other in each case (AML and ALL).

How to Answer Question k)

Remember that “high” corresponds to “more red” while “low” corresponds to “more blue”. In this problem we’re interested in the TYPE of tumor (ALL or AML). Use the expressions of the other samples (columns) to identify which type this new sample most likely belongs to.