Program that gives me protein similarity

Question

[andre.arrudalima]

Hi,

I have some protein alignments and want to find their similarities. I tried using Uniprot, but it only provides identity information. I also tried using MEGA, but couldn't find a way to access similarities there.

Can someone help me find the similarities, please? Preferably in MEGA, but I would be grateful for any online tool as well.

Thank you.

[GenoMax]

Can you provide a specific example of what you are looking to get? Are you looking for residues that are able to replace amino acids without changing function/structure of a protein?

[andre.arrudalima]

I'm not trying to get in this level of detail. I only want to compare aminoacid for their chemical characteristics, which I think is the default for a similarity comparison.

Anyway, here are my sequences, the kinase domain of some plant receptors

> FER 
FDESRVLGVG GFGKVYRGEI DGGTTKVAIK RGNPMSEQGV HEFQTEIEML SKLRHRHLVS
LIGYCEENCE MILVYDYMAH GTMREHLYKT QNPSLPWKQR LEICIGAARG LHYLHTGAKH
TIIHRDVKTT NILLDEKWVA KVSDFGLSKT GPTLDHTHVS TVVKGSFGYL DPEYFRRQQL
TEKSDVYSFG VVLFEALCAR PALNPTLAKE QVSLAEWAPY CYKKGMLDQI VDPYLKGKIT
PECFKKFAET AMKCVLDQGI ERPSMGDVLW NLEFA

> THE1 
FDESSLLGVG GFGRVYKGTL EDGTKVAVKR GNPRSEQGMA EFRTEIEMLS KLRHRHLVSL
IGYCDERSEM ILVYEYMANG PLRSHLYGAD LPPLSWKQRL EICIGAARGL HYLHTGASQS
IIHRDVKTTN ILLDENLVAK VADFGLSKTG PSLDQTHVST AVKGSFGYLD PEYFRRQQLT
EKSDVYSFGV VLMEVLCCRP ALNPVLPREQ VNIAEWAMAW QKKGLLDQIM DSNLTGKVNP
ASLKKFGETA EKCLAEYGVD RPSMGDVLWN LEYA

> AT5G01950 
FSSSTLVGRG GYGKVYRGVL SDNTVAAIKR ADEGSLQGEK EFLNEIELLS RLHHRNLVSL
IGYCDEESEQ MLVYEFMSNG TLRDWLSAKG KESLSFGMRI RVALGAAKGI LYLHTEANPP
VFHRDIKASN ILLDPNFNAK VADFGLSRLA PVLEDEEDVP KHVSTVVRGT PGYLDPEYFL
THKLTDKSDV YSIGVVFLEL LTGMHAISHG KNIVREVKTA EQRDMMVSLI DKRMEPWSME
SVEKFAALAL RCSHDSPEMR PGMAEVVKEL ESLL

> BAK1 
FSNKNILGRG GFGKVYKGRL ADGTLVAVKR LKEERTQGGE LQFQTEVEMI SMAVHRNLLR
LRGFCMTPTE RLLVYPYMAN GSVASCLRER PESQPPLDWP KRQRIALGSA RGLAYLHDHC
DPKIIHRDVK AANILLDEEF EAVVGDFGLA KLMDYKDTHV TTAVRGTIGH IAPEYLSTGK
SSEKTDVFGY GVMLLELITG QRAFDLARLA NDDDVMLLDW VKGLLKEKKL EALVDVDLQG
NYKDEEVEQL IQVALLCTQS SPMERPKMSE VVRMLEGDGL AERWEEWQ

>FLS2 
FNSANIIGSS SLSTVYKGQL EDGTVIAVKV LNLKEFSAES DKWFYTEAKT LSQLKHRNLV
KILGFAWESG KTKALVLPFM ENGNLEDTIH GSAAPIGSLL EKIDLCVHIA SGIDYLHSGY
GFPIVHCDLK PANILLDSDR VAHVSDFGTA RILGFREDGS TTASTSAFEG TIGYLAPEFA
YMRKVTTKAD VFSFGIIMME LMTKQRPTSL NDEDSQDMTL RQLVEKSIGN GRKGMVRVLD
MELGDSIVSL KQEEAIEDFL KLCLFCTSSR PEDRPDMNEI LTHLMK

[GenoMax]

I see. Have you tried a multiple sequence alignment in MEGA? I don't have MEGA available immediately but the color coding options should allow you to look at chemical similarity, if I recall.

[andre.arrudalima]

Yes, I did. The point is there I want to see a number cause I want to make a similarity table. I cant find similarity or identity percentages on MEGA.

Answer 1

[Sajad]

try blastp. if you have a bulk of sequences you can use local blast.

[andre.arrudalima]

Thank you. Doing this, the blast only give me indentities. Is the "Positive" the similarity?

[GenoMax]

Yes. That is correct.

From: https://www.ncbi.nlm.nih.gov/books/NBK62051/

substitution : The presence of a non-identical amino acid at a given position in an alignment. If the aligned residues have similar physico-chemical properties or have a positive score in the governing scoring matrix the substitution is said to be conservative.

[andre.arrudalima]

Hi again I just realize that blasp only gives me the similarity of a specific part o the protein that it align, overestimating the global similary. For example, I have a 100 aa protein that align just 70aa to another protein, but with 100% similarity. Blastp will give me 100% similarity but in reality the real complete aligment should give me 70%. How can I solve this?

[GenoMax]

BLAST is a local alignment algorithm (as the name indicates). If you need to get global estimates you will need to use a global alignment algorithm like Needleman-Wunsch.

[andre.arrudalima]

Thank you for the answer and sorry for the multiple post. I though you wouldn see here.

I have two questions: 1) Is it possible to loose blast algorithms in a way that it will align the whole protein? 2) Is there a online tool with Needleman-Wunsch algorithms? I not willing to do a data base search, just a small aligment with less than 15 proteins.

[GenoMax]

No for the first question. For second you can use emboss needle: https://www.ebi.ac.uk/Tools/psa/emboss_needle/