Invisible Illness Awareness Week 2016: Our Voices Need to Be Heard
Never heard of Invisible Illness Awareness Week? You're not alone. Jody Smith sheds a little light to make it more visible
Discuss the article on the Forums.

Predicting Rates (and inferring Severity) of Mutations

Discussion in 'Genetic Testing and SNPs' started by Valentijn, Oct 28, 2014.

  1. Valentijn

    Valentijn Senior Member

    Messages:
    14,281
    Likes:
    45,823
    I'm taking a Bioinformatics class in Coursera (generally not interesting unless you REALLY want to do mental gymnastics), and one thing we learned about recently is that predictions are made regarding the effects of amino acid substitutions in proteins.

    Basically when there is a different variation in a SNP in a gene (such as having an "A" allele instead of "G"), sometimes it results in a missense mutation in the protein created by the gene. That just means that one amino acid, such as cysteine, is substituted with another amino acid, such as glycine.

    That can result in the protein behaving much differently, but often has little or no impact. But by looking observing the actual rates of one amino acid being replaced with another (BLOSUM), or by calculating the likelihood of it happening based on the similarities and differences between amino acids (PAM), various matrices (charts) have been created which show the likelihood of specific amino acid mutations occurring.

    This also has implications regarding how much effect a mutation is likely to have. If it's a very common one, such as from leucine to isoleucine, it probably won't have much impact due to those two amino acids being extremely similar, and accordingly it'll be fairly common to see such mutations. Whereas glycine behaves very differently than either leucine or isoleucine, and a mutation from glycine to one of those would be much more uncommon.

    Here's one such commonly used BLOSUM matrix:
    [​IMG]
    Big numbers mean the mutation is more common, and the most negative numbers mean the mutation is least common.

    And here's a chart showing the shared amino acid characteristics used in predicting values for PAM matrices:
    [​IMG]

    Anyhow, it could be relevant in determining how interesting some of our own mutations are :D
     
    SOC, Critterina, ahmo and 1 other person like this.
  2. taniaaust1

    taniaaust1 Senior Member

    Messages:
    11,916
    Likes:
    12,773
    Sth Australia
    Sounds very interesting, I wish I could understand that stuff but its too much for my mind to even slightly start to understand it. I know Im going to ask you some questions about this thou (not tonight as I only just been released from hospital again so need to relax tonight).
     
    Valentijn likes this.
  3. Critterina

    Critterina Senior Member

    Messages:
    1,202
    Likes:
    700
    Arizona, USA
    What do the numbers on the diagonal mean? It seems that they are not substitutions at all.
     
    Valentijn likes this.
  4. Sea

    Sea Senior Member

    Messages:
    1,268
    Likes:
    2,650
    NSW Australia
    Just guessing that they're the synonymous mutations where the snp substitution doesn't change the protein
     
    Valentijn likes this.
  5. Valentijn

    Valentijn Senior Member

    Messages:
    14,281
    Likes:
    45,823
    Yup, and the ones with the biggest numbers there seem to be the ones which are the least likely to change into something else
     

See more popular forum discussions.

Share This Page