if I do a blast search using my amino acid sequence to look for homologous proteins and then I use my nucleotide sequence will I get the same results? From talking to colleagues the answer is no but why? Is one better than the other in certain circumstances? I can see the fact that several open reading frames can exist on a nucleotide sequence and that may "pollute" the results- but could this be beneficial in any way? I haven't tried this yet but I thought it might be interesting to get people's comments on this.
Thanks?
Because there are only 20 amino acids involved in protein synthesis, but 64 possible codons, there is a variable degree of degeneracy. For example, although Met and Trp are encoded by unique codons, Arg, Ser and
Leu each have six. Given different GC contents and codon preferences in different organisms, you may have a much lower level of sequence identity at the nucleotide level than at the protein level. Thus, when looking for related sequences in the database, a protein search should be more sensitive than a nucleotide search. Obviously, this only applies for coding regions: for non-coding regions, the selection pressure maintaining sequence homologies acts at the level of the nucleotide
sequence.
Because there are only 20 amino acids involved in protein synthesis, but 64 possible codons, there is a variable degree of degeneracy. For example, although Met and Trp are encoded by unique codons, Arg, Ser and
Leu each have six. Given different GC contents and codon preferences in different organisms, you may have a much lower level of sequence identity at the nucleotide level than at the protein level. Thus, when looking for related sequences in the database, a protein search should be more sensitive than a nucleotide search. Obviously, this only applies for coding regions: for non-coding regions, the selection pressure maintaining sequence homologies acts at the level of the nucleotide
sequence.
Completely agree with Atif Idrees. To be more simple, protein level is functional, altough mRNA are more accurate than genes because they are already spliced, you may have a lower score in homology alignments on mRNAs and the coded proteins may be instaed functinally related and more similar at amminoacidic level homology.
With genes sequences you may also have very similar proteins coming from low homology genes.
You may have enormously different keys in shape, color and size, but in order to open the same door with different keys the part that must enter the door hole should be the same, fit and sit ( because it's "funcional" ).
The benefits of using protein sequences over DNA sequences (comparing coding regions):
a.) 20 amino acids instead of 4 Bases and a shorter sequence input increase the sensitivity of your results significantly
b.) Like pointed out before, the degenerated DNA code may led to “noise” (not significant results or decreased sensitivity due to silent mutation)
c.) By using a protein weight matrix we can also take physico-chemical properties into consideration (some amino acid substitutions are more likely to occur then others)
d.) DNA may include non-coding regions leading to false results (depends on your data of course)
for more detailed information I suggest: http://www.ncbi.nlm.nih.gov/books/NBK20261/
Thank you for sharing your knowledge and experience with me. That was very kind of you.
- Badges
- Science method
- Similar topics
- Biological Science
- Genetics
- Genetic Analysis
- genetic techniques
- Sequencing