Please enable JavaScript.

Coggle requires JavaScript to display documents.

Alignment, ALIGNMENT - Coggle Diagram

- - - - Compute for each base in a sequence using 2d matrix
      - Trace back
      - Lecture 7, slide 18
      - Initialize first row and column with gap costs
  - - - To identify highly conserved DNA regions involved in basic biological functions
      - To identify protein domains (modules re-used by different proteins)
    - - Algorithm consists of 0 as its minimum, not allowing negative scores
      - Initialize first row and column with 0s
      - Different cost for the unaligned ends
    - - Few small queries can be searched for in many large database sequences
      - Seeding - First finds high scoring matches between query Kmers (exact and similar) (length 11 for DNA and 3 for protein) and database sequences.
        
        Then extends local alignment in both directions to find best match around the seeds.
      - E value - expected number of hits with at least this score, against any database of this size.
    - - Cost - to minimise
        
        Simple costs
        
        All insertions and deletions are weighted same
        
        Affine gap costs
        
        Greater cost to open a gap than to maintain one
        
        Results in lesser matches, but biologically more valid
      - Score - to maximise
      - More realistic if dependent on substitution matrix - E.g. BLOSUM, PAM, etc. (for protein)
  - - - Many reads against few database sequences - opposite of BLAST - so it is not suitable
      - Tools
        
        BWA, Bowtie, BFAST, SHRiMP, Kanga, SAMtools
        
        For read mapping, BWA (Burrow Wheeler alignment)
        
        Build an index/hash table of reference
        
        Find good potential matches between reads and index - where in the reference does the read match the longest
        
        Perform semi-global around the matches
        
        Can also perform local alignment (clipping of reads) around matches
      - Formats - SAM/BAM
        
        Where read has aligned, how much of it is aligned, how well it has aligned, difference with reference, etc.
      - Reference Alignment of reads - Read Mapping
        
        chrUn - chromosome unknown in database
        
        Assigned contigs in the database whose chromosomal location is unknows
        
        Mainly there to increase read mapping accuracy and decrease false positive variant calls
      - Doesn't penalize gaps at either end of alignment
    - - Reasons for requirement of inexact alignment
        
        Reference and sample are different
        
        Poor reference assembly
        
        Different organisms being compared
        
        Individual differences within organisms
        
        Cancer Vs normal
        
        Errors during sequencing