The following exercises will illustrate several aspects of pairwise sequence comparison:

• Using some important web resources for sequence analysis software, presented in the preceding practicals.
  

  Swiss EMBnet Node	This server gives access to several sequence comparison tools. Use this link as entry point for the following practicals !
  ExPASy Server	Gives access to the protein database SWISS-PROT and other usefull proteomics tools.
  NCBI Server	Gives access to the GenBank/EMBL/DDBJ DNA sequence database and several other databases and tools.

• Understanding the usage of web based sequence analysis tools.
• Using pairwise sequence comparison tools to illustrate some theoretical aspects studied during the course.

Reminder: some links for the recovery of sequences

[ Easily fetch sequences ] [ Search SWISS-PROT ] [ NCBI's Entrez ]

Compare the sequences OPRM_RAT and SSR1_HUMAN (these are the SWISS-PROT IDs) with lalign using default parameters.

The sequences can be fetched here (choose the "FASTA" format) using the SWISS-PROT IDs.
Don't hesitate to look at the complete SWISS-PROT entries (OPRM_RAT and SSR1_HUMAN), in order to get more information about these two proteins !

Try to answer the following questions:

• Is this a local or global alignment ?
• Switch between local and global alignment . Try to understand the differences.
• Why are there several alignments displayed when performing the local alignment ?
• What does "% identity" mean ? How is it computed ?
• What do the symbols ":" and "." stand for ?
• When two residues are different, there can be either a "." or a blank. Try to understand the difference and what parameters influence this result ?
• Try to modify the gap penalties, examine more closely how these parameters influence the occurence and the length of gaps ("-").
• Try to modify the scoring matrices used (i.e. BLOSUM35 and BLOSUM80), examine more closely how these parameters influence the scores and the alignments.

Compare the same sequences (OPRM_RAT and SSR1_HUMAN) using Dotlet (If you are working on a Mac Dotlet may not work).

The sequences can be fetched here (choose the "FASTA" format).

Start with a look onto the Dotlet documentation

• Load the two sequences into Dotlet and compute the dotplot.
• What does the intensity (gray level) of a pixel mean ?
• Try changing the grayscale borders. Where would be an optimal position for the upper and lower limits of the grayscale ?
• What do the diagonal lines represent ?
• Try to identify corresponding aligned regions in the dotplot and the alignments found by LALIGN.
• Try to modify the noise by changing the window size, the threshold, both.
• Try comparing each sequence against itself.

The Dotlet learn by example pages show different typical sequence analysis problems.

• Take an interested look at the Dotlet examples.
• Try to understand the dotplots.

For those who can't get enough: get some more practice by compairing some of the pairs of sequences below.

• CO9_HUMAN - PERF_HUMAN
• FOSL_DROME - GCN4_YEAST
• HBB_HUMAN - LGB1_PEA
• YOR6_ADEG1 - CD4_HUMAN