Practical - Multiple Alignment & Phylogeny

Part 1 - Analysis of a Protein Family

A protein sequence (available on the Moodle platform) has been automatically predicted during the annotation of the fungus Malassezia globosa genome.

Similarity Search

Perform a similarity search using BLAST at EBI in the SWISSPROT database (https://www.ebi.ac.uk/jdispatcher/sss/ncbiblast) (default parameters).

Exercise 1 How many sequences do you obtain (with E<10)? Why so few sequences are detected according to you?

Display results with Tool output.

Exercise 2 Find 3 conserved motifs.

Exercise 3 Is the Malassezia sequence homologous to: Q84424, P05469, P0C995, O13881?

Display results with Visual output.

Exercise 4 For each group of homologs, specify which region(s) (N-terminus, central region, C-terminus, almost whole sequence…) of the query sequence and of the database sequences have been aligned by BLAST.

Exercise 5 Why can this information be important before constructing a global multiple alignment?

Display the Summary table. Select all sequences detected by BLAST except:

• CLR2_SCHPO (O13881) and YKP3_KLULA (P05469): non-homologous sequences
  
• YL226_MIMIV (Q5UQC0) and MCE_MIMIV (Q5UQX1): too few motifs.

Save sequences in Fasta format and add the query sequence in the file. The final file is also available on Moodle (distant_homologs.txt).

Multiple Alignment of Homologs Using Mafft

Parameters:

Display alignment parameters.

Exercise 6 Why are pairwise parameters mentioned?

Align sequences (Alignment, Do complete alignment) with default parameters.

Visualize this alignment with Jalview (available on your computer or using https://www.jalview.org/jalview-js/JalviewJS/).

Residues can be colored according to their properties with Colour, Zappo (or Clustal, for instance).

Exercise 7 Is the sequence order relevant? Modify sequence order if necessary.

Exercise 8 Do all sequences exhibit the same domain organization?

Exercise 9 Do you find the previously defined motifs? According to you, is the alignment correct?

Exercise 10 Can you find additional motifs? Can you find obvious sequence prediction errors?

Exercise 11 Which sequence sub-groups can be defined? Are there “orphan” sequences? Justify your answer and give one discriminative residue or motif for each group.

Phylogeny

A multiple alignment of aspartyl-tRNA synthetases (SYD) and asparaginyl-tRNA synthetases (SYN) protein sequences has been constructed. A few lysyl-tRNA synthetase (SYK) sequences have been added as an outgroup. In Eukarya, cytoplasmic proteins are indicated with C (ex: SYNC_HUMAN) while mitochondrial proteins are indicated with M. The alignment is schematically shown below (Jalview overview).

Exercise 12 According to this graphical overview, what precaution should we take before building a phylogenetic tree from this alignment?

Some sequences have been selected, and the variable N-terminus part of the alignment has been removed. The corresponding alignment is available on Moodle (fasta format).

Codes for species:

• Animals: human, mouse.
  
• Fungi: yeast, Schpo, Enccu.
  
• Amoebas: dicdi.
  
• Archaea: pyrab, aerpe
• Bacteria: bacsu, ecoli
• Plant: arath

Use the Seaview program available on your computer (http://doua.prabi.fr/software/seaview) and open the multiple alignment with Seaview.

Distance Method (BioNJ) – Pairwise Gap Removal

Construct a phylogenetic tree using BioNJ (Trees/Distance Methods/) with 100 bootstrap replicates, Poisson distance, and ignore all gaps (global gap removal).

Exercise 13 How many sites have been used to calculate distances? Do you agree with the root position?

Exercise 14 Which pair of sequences are closest to each other? Which sequence has the highest rate of evolution (except SYK_BACSU)?

Exercise 15 Display bootstrap values using “Br support”. Is the separation between aspartyl-tRNA synthetases (SYD) and asparaginyl-tRNA synthetases (SYN) sequences clearly established according to this tree?

Exercise 16 Are the relationships between SYN sequences expected? Are they significantly supported?

Exercise 17 Are phylogenetic relationships observed for eukaryotic aspartyl-tRNA synthetases in agreement with the current phylogeny of eukaryotes?

Exercise 18 What is the homology relationship between:

• sync_human and sync_mouse?
• sync_human and sydc_mouse?
• sync1_arath and sync2_arath?

Exercise 19 Why is it important to distinguish these types of relationships?

Distance Method (BioNJ) – Pairwise Gap Removal

Construct the phylogenetic tree using BioNJ (Trees/Distance Methods/) with 100 bootstrap replicates, Poisson distance but without the “ignore all gaps” option (=>pairwise gap removal).

Exercise 20 How many sites are used on average?

Exercise 21 Are there topology differences between the two trees? Apart from topology, what types of differences can be observed?

Maximum Parsimony (ProtPars)

Construct the phylogenetic tree using Maximum parsimony (default parameters but with 100 bootstrap replicates).

Exercise 22 What can you say about the execution time of the two programs?

Exercise 23 How many sites are informative? What do “steps” stand for? How can we obtain “3 best trees”?

Exercise 24 Is branch length informative in this tree?

Exercise 25 The tree is shown as a rooted tree. Do you agree with the root position? Modify the root position if needed (“re-root”).

Exercise 26 What is the most reliable tree between the first tree and this one? Justify your answer.

Maximum Likelihood (PhyML)

Construct the phylogenetic tree using PhyML: default parameters but without calculating branch support (“none”).

Exercise 27 Is the tree topology in agreement with the topology obtained using BioNJ?