Protein function is dependent on protein 3D structure more than its sequence. Analyze the structure of a protein obtained from mutated sequence against wild sequence. You could try sites like UNIPROT KB... Also analyse whether the mutations are in the domain or conserved part of the sequence...

Hi there,

Prediction of mutation effect on protein biological activity is a tricky guess... Especially when you don't have any structural data. Then first try to identify sequence homologs for which 3D structures are known and try to generate a structural model of your protein. The model should help you to evaluate mutation effect on structure and activity (if you have identify the active site). But there is nothing like experiment to confirm...

Hi Monica, I'm following your question because I'm very interested in knowing if there is a database out there that can predict protein function!

Anyway, as Danish pointed out, the effects of a mutation in protein function depends on where the mutation is.

It would help if you could give more information about the protein in question. Is it an enzyme, structural protein, transcription factor, etc.

If it is a described protein, you can do a BLAST of the sequence to see what mutations have been documented and what their effect has on function. 

You may find this link interesting, they have some documents and explanations, there have a challenge where many groups are trying to do the same.

http://www.predictioncenter.org

Hi Monika,

As stated by many above, finding the "best" method/answer for a particular system is case dependent. One would need more details, such as type of protein and  what induced the mutations.  Often studies of such things involve use of some kind of drugs as probes to map a specific function of genetic change . Below are three papers that provide examples using different types proteins with differing levels of structural/biological data available to discuss effect of mutation on form/function. Feel free to write back for more details as needed.

1. Nettles, J. H.; Stanton, R. A.; Broyde, J.; Amblard, F.; Zhang, H.; Zhou, L.; Shi, J.; McBryer, T.; Whitaker, T.; Coats, S.; Kohler, J.; Schinazi, R. Asymmetric Binding of BMS-790052 to HCV NS5A Suggests Two Novel Modes of Action. Journal of Medicinal Chemistry In Press dx.doi.org/10.1021/jm501291c.

2. Schinazi, R. F.; Massud, I.; Rapp, K. L.; Cristiano, M.; Detorio, M. A.; Stanton, R. A.; Bennett, M. A.; Kierlin-Duncan, M.; Lennerstrand, J.; Nettles, J. H. Selection and characterization of HIV-1 with a novel S68 deletion in reverse transcriptase. Antimicrob Agents Chemother 2011, 55, 2054-60.

3. Hari, M.; Loganzo, F.; Annable, T.; Tan, X. Z.; Musto, S.; Morilla, D. B.; Nettles, J. H.; Snyder, J. P.; Greenberger, L. M. Paclitaxel-resistant cells have a mutation in the paclitaxel-binding region of beta-tubulin (AsP(26)Glu) and less stable microtubules. Molecular Cancer Therapeutics 2006, 5, 270-278.

Dear Monika

You have already received some very good answers.

One thing to note: you ask "how this mutation affects the protein function?". Well, maybe is does and maybe it doesn't. It depends on a number of factors, the nature of the amino-acid switch, where it is located in the sequence, what it's neighbours are in the sequence, and most important (as one of your responders has pointed out) which contacts in 3D are altered by substituting one amino acid residue for another.

Since you have this sequence database you should begin by making a multiple sequence alignment using MaxHom or more modern methods and then look at what the various replacements are at the key sites in your sequence. Look at the mutations at the residue position of interest and then look for correlated mutations i.e. other sites where changes also take place with the same kind of pattern as you observe at the first site. That is usually indicative of some kind of 3D contact.

You haven't specified what function your are interested in, however. Most proteins have many functions ....

Dear all, 

thank you very much for all your answers, they are very helpful! 

I have no knowledge of the protein function so far, and there isn't anything in databases connected to it - it's a zebrafish protein, homologous to a human dyslexia candidate. What I'd like to have is a knock-out zebrafish of that protein. I have found a few mutations listed in Sanger database, but I don't know what they mean. I believe the one with disrupted Essential Splice Site would be a good candidate to look at then.The only other option I have is the one with a nonsense mutation.

It sounds like you are looking at DNA sequences. In DNA mutations can be silent or not - but you clearly already know that.

If you can identify the ORF, then translate that sequence to protein sequence. Take the (single letter code) protein sequence and paste it into you favoúrite sequence alignment tool. My favourite is PredictProtein. You will get out a huge amount of useful information - alignments, alignment statistics (like variability, entropy etc), PDB structures etc etc.

If you want to take that alignment and do correlated mutation studies, I can tell you how to do that, but run the PredictProtein first.

Hi Monika,

I made some similar work a long time ago on BRCA1

http://www.ncbi.nlm.nih.gov/pubmed/?term=huyton+BRCA1

If you know some structural information (i.e have a crystal structure) then there are lots of things you can do.

amino acid substitutions can be "modelled" and delta delaG values calculated to estimate the difference in stability of the mutant protein

http://mordred.bioc.cam.ac.uk/~sdm/sdm.php

http://folding.biofold.org/i-mutant/i-mutant2.0.html

http://foldx.crg.es/

http://www.hsls.pitt.edu/obrc/index.php?page=URL1258139151

can all do this

st andrews is full of structural biologists, i am sure someone will be able to help you out.

Hi,

Protein function can be correlated with conformation of the protein or protein structure. First you model the protein structure and then observer the mutational changes in your structure via checking the stability of the protein i.e validating the protein in servers like Prosa, Procheck, etc

Danish J

There are many online servers like SIFT, PolyPhen2, SDM etc. which predict the effect of amino acid mutation. These servers are based on different algorithms so it would be better if you have a consensus result.

The most deleterious mutations are those that result in a premature stop codon: the earlier in the sequence it occurs, the more likely you will find a total loss of function.

Look at an amino acid sequence alignment of homologs (preferably orthologs) over a wide variety of species: Mutations in positions that are highly variable are less likely to have an effect on function that mutations that lie in highly conserved positions. Mutations from (conserved) Gly to non-Gly residues and mutations to Pro are most likely to affect structure, since these reduce the range of acceptable main-chain torsion angles. For extracellular proteins, Cys mutations are also likely to have an effect: loss of a disulfide bond may destabilize your protein, while unpaired Cys residues can lead to crosslinking if the side chain is accessible.

Conservative substitutions are likely to have a milder effect than non-conservative ones. Here too, a multi-sequence alignment can give you clues on what property is important: size, charge , hydrophobicity, hydrogen bonding.

You can use missence3d