Point Accepted Mutation Matrix
This article has been registered in the quality assurance biology for improvement due to formal or content-related deficiencies . This is done in order to bring the quality of the biology articles to an acceptable level. Please help improve this article! Articles that are not significantly improved can be deleted if necessary.
Read the more detailed information in the minimum requirements for biology articles .
A Point Accepted Mutation Matrix ( PAM matrix for short ) is a substitution matrix filled with PAM values that is used in bioinformatics to determine the probability of a change in a nucleotide sequence . The basis for creating a PAM matrix are statistically recorded values based on sequence differences.
In practice, the PAM250 matrix is the most commonly used. It shows a sequence match of about 20%, which means that a correct alignment can be expected from the sequence analysis alone . It was first described by Margaret Oakley Dayhoff , who published it in 1978 in the "Atlas of Protein Sequence and Structure".
Motivation & creation
The further sequences develop, the more mutations accumulate. Margaret Oakley Dayhoff was the first to try to determine which amino acids would mutate into a different amino acid with what probability. The basis for this observation is that the chemical properties of an amino acid residue usually remain untouched, an acidic amino acid is thus replaced with a comparatively high probability against a likewise acidic.
An important factor regarding the use of these observed change probabilities is the question of how much two sequences generally differ. The sequence
AGSTVAVREA
| |||||| |
ATSTVAVRSA
has a distance of 2 and thus a difference of 20%. Depending on how much the alignments of two or more sequences match, a distinction is made between 1-PAM matrices (corresponds to 1% accepted mutation) and 250-PAM matrices (corresponds to around 80% accepted mutations).
The entries in the PAM matrix are given as log-odds values and multiplied by a certain factor (usually 10) in order to avoid decimal numbers . To close by a score of the matrix to the mutation probability, the value is divided by 10 and then to the base 10 potentiated . A value of +2 in a PAM matrix means that amino acid A mutates to amino acid B 1.6 times more often than expected (2/10 = 0.2 and 10 ^ (0.2) = 1.6).
In order to obtain reliable results even for sequences with very little similarity, S. Henikoff and JG Henikoff developed the family of BLOSUM matrices.
literature
- http://helix.mcmaster.ca/721/distance/node9.html ( Memento from November 13, 2007 in the Internet Archive )
- Arthur M. Lesk: Bioinformatics - An Introduction . Heidelberg; Berlin: Spektrum, Akad. Verlag, 2003, ISBN 3-8274-1371-0
Individual evidence
- ↑ MO Dayhoff, R. Schwartz, BC Orcutt: A model of Evolutionary Change in Proteins. In: Atlas of protein sequence and structure , 5th edition, 3rd supplementary volume, 1978, Nat. Biomed. Res. Found., ISBN 0-912466-07-3 , pp. 345-358