Introduction to Unix for Bioinformatics

Handling genomic coordinates


Opening a session.

Starting a terminal

  1. Connect to the computer using your login and password.
  2. Open a terminal. You should find a terminal (terminal, konsole, terminator,...) under the Application Menu: Applications > Accessories > Terminal

Using this terminal you should now be able to communicate with the system through a set of shell commands (your default shell command interpreter should be BASH).

Creating a working directory

  1. Using the mkdir command (make directory) create a working directory named td1_genomic_coordinates in your HOME folder.
  2. Use the cd (change directory) command to enter the td1_genomic_coordinates directory
View solution| Hide solution [back to contents]

Downloading RefSeq transcript coordinates

RefSeq transcript

The Reference Sequence (RefSeq) collection aims to provide a comprehensive, integrated, non-redundant, well-annotated set of sequences. Here we will download informations about RefSeq transcripts (that is known full-length transcript) related to Mus musculus genome (mm9 version). Refseq transcript can be obtained from UCSC genome browser.

  1. Go to the UCSC web site
  2. In the middle of the left menu select "Downloads".
  3. Select the Mouse genome.
  4. In the "Mar. 2006 (mm9)" section select "Annotation database" to enter the UCSC ftp web site.
  5. Find the file refGene.txt.gz file copy the link location (URL) by right-clicking on the link.
  6. In a terminal use the wget command followed by the address to retrieve the refGene.txt.gz file.
  7. Uncompress the file using the gunzip command.
View solution| Hide solution

What kind of informations are available for each transcript?

  1. Go to the UCSC web site
  2. Select Tables in the top menu
  3. Select Clade: Mammal, Genome: Mouse, assembly: mm9, group: Genes and Gene Prediction tracks, table: refGene and click on describe table schema

Let's have a look at the refGene file?

  1. Use the wc command with -l arguments to count the number of lines in the refGene.txt file
  2. Use the head command with -n arguments to see the first 5 lines of the refGene.txt file.
  3. Use the tail command with -n arguments to see the last 5 lines of the refGene.txt file.
  4. Have a look at the whole file using less (use up/down arrow keys and type "q" to quit))
  5. Using head extract the first line and use a pipe to redirect the output to the od command with -c argument.

What can you guess from the results?

View solution| Hide solution

Some descriptive statistics on refSeq transcript using basic unix command

  1. Using the cut command retrieve the 13th column
  2. Using the cut command retrieve the 13th column and redirect the output to the grep command to find genes whose symbol start with Bcl. Count them by redirecting the output to the wc -l command
  3. What is the gene that display the highest number of transcripts (use sort and uniq with -c argument). Check the results by typing the gene symbol in the UCSC genome browser (top menu > genomes)
  4. What is the gene that display the highest number of exons.
  5. Some transcrits may be coding (NM_...) or non-coding (NR_...). How many transcripts of each class are available in the refGene.txt file?
  6. How many genes does each chromosome holds?
View solution| Hide solution

The Bedtools suite

What is bedtools

The bedtools manual says:

The BEDTools utilities allow one to address common genomics tasks such as finding feature overlaps and computing coverage. The utilities are largely based on four widely-used file formats: BED, GFF/GTF, VCF, and SAM/BAM. Using BEDTools, one can develop sophisticated pipelines that answer complicated research questions by "streaming" several BEDTools together. The following are examples of common questions that one can address with BEDTools.
  1. Intersecting two BED files in search of overlapping features.
  2. Computing coverage for BAM alignments based on genome features.
  3. Merging overlapping features.
  4. Screening for paired-end (PE) overlaps between PE sequences and existing genomic features.
  5. Calculating the depth and breadth of sequence coverage across defined "windows" in a genome.
  6. Screening for overlaps between "split" alignments and genomic features.

Installing the bedtools suite

Check if the bedtools suite is already installed on your system (type intersectBed in a terminal). If bedtools is not available, follow the procedure below to perform installation.

      mkdir -p  ~/bin
      mv BEDTools.v2.16.2.tar.gz ~/bin
      cd ~/bin
      tar xvfz BEDTools.v2.16.2.tar.gz
      cd BEDTools-Version-2.16.2
      echo PATH=$PATH:~/bin/BEDTools-Version-2.16.2/bin >> ~/.bashrc
      source ~/.bashrc

Transforming refGene file into bed format

First we need to transform the refGene.txt file into a bed to use it as input for the bedtools suite. We will do it with awk.

      awk 'BEGIN{FS=OFS="\t"}{print $3,$5,$6,$2"|"$13,0,$4}' refGene.txt> refGene.bed

Upload the resulting bed file into UCSC (top menu, genome > add custom track > choose file > submit). Go to the genome browser to visualize the new track.

Getting promoter regions of each transcript

To obtain promoter regions of each transcript we will first extract the TSS coordinates of each transcript. The TSS will correspond to the txStart or txEnd coordinates depending on wether genes is located on the positive or negative strand respectively ( see refSeq table definition).

      awk 'BEGIN{FS=OFS="\t"}($6=="+"){print $1,$2,$2,$4,$5,$6}' refGene.bed > refGene.TSS.bed
      awk 'BEGIN{FS=OFS="\t"}($6=="-"){print $1,$3,$3,$4,$5,$6}' refGene.bed >> refGene.TSS.bed

Now we will use slopBed to increase the size of each feature (TSS) to define a promoter region around the TSS ([-2000,+500]). Also this can be done with a simple awk command, sploBed will restrict the resizing to the size of the chromosome (i.e. no start < 0 and no end > chromosome size). We thus need to download the length of each chromosome (here from UCSC ftp site).

      gunzip chromInfo.txt.gz
      cat chromInfo.txt

Now we can run slopBed. We will also reorder features (TSS) according to chromosomes coordinates

      slopBed -i refGene.TSS.bed -l 2000 -r 500 -s  -g chromInfo.txt  | sortBed > refGene_Prom_2000_500.bed

Check the results by loading the refGene_Prom_2000_500.bed as a new user track in the UCSC genome browser.

NB: The RSAT web server also proposes advance tools for retrieving promoter regions (excluding coding regions, repeat elements,...). This tools can also be run using UNIX command lines.

Intersecting promoter regions with predicted transcription factor binding sites

The dataset that will be used thereafter was obtained from the following publication:

In this article, the authors scanned vertevrates alignments in search for highly conserved putative transcription factor binding sites (TFBS). The resulting file contains the coordinates of the predicted TFBS (mm9 genome version)

  1. Download the conserved_predicted_TFBS_mm9.bed.gz file
  2. Use intersectBed to find for each promoter region its associated sets of predicted TFBS.
  3. Using grep and sed select the non-redondant list of targets for transcription factor E2F (represented by M00920:V$E2F_Q6_01 identifier), or ETS (represented by M00771:V$ETS_Q4 identifier), or MEF2A (represented by M00941:V$MEF2_Q6_01 identifier), or NFKB (represented by M00054:V$NFKAPPAB_01 identifier) or NRSF (represented by M01028:V$NRSF_Q4 identifier).
View solution| Hide solution

Functional annotation using Database for Annotation, Visualization and Integrated Discovery

We have seen that each transcription factor is associated with a set of putative targets. What can we learn from these targets? Is the transcription factor associated to genes involved in some particular biological functions?

To answer this question we will use the Database for Annotation, Visualization and Integrated Discovery (DAVID). DAVID will try to find whether the selected genes share some particular features, by analyzing their annotations from several sources including Gene Ontology, KEGG pathways, Reactome Pathways, genomic locations, ...

  1. Go to the DAVID web site.
  2. In the left menu select Functional annotation.
  3. In the left menu select the upload tab. Paste the list of genes associated to one of the transcription factors.
  4. Select step 2 > OFFICIAL_GENE_SYMBOL.
  5. Select step 3 > Gene List.
  6. Press Submit List.
  7. The program now displays the "Gene list manager", with a list of species sorted by decreasing number of matching gene names. Select "Mus musculus" and click "Select species".
  8. Optionally, you can click on the list under "List manager" and rename if (for example, name it "M00920 E2F mouse").
  9. In the "Annotation Summary Results" page, click the button Functional Annotation Chart.
  10. In the new window, select option and ask for Fold Enrichment, Bonferroni, Benjamini, FDR, Fisher Exact, LT,PH,PT

Note: the Fisher test will be explained in the next practical "GO statistics".

What can you guess from the results?

  1. Which are the most significant functional classes?
  2. How do you interpret their P-value?
  3. Is the significance proportional to the fold enrichment?
  4. Compare the P-values obtained after correction for multiple testing (Boneroni, Benjamini).
  5. Evaluate the results at the bottom of the table: can these associations be considered as significant? Why?
View solution| Hide solution