If you are only interested in single protein, then blasting it against NR database is the best solution. You can customize the blast to output taxon id like the following: -outfmt '6 qseqid sseqid evalue pident qcovs sgi sacc staxid ssciname scomname stitle'

Once you have the taxon ids/stitle, you can use the taxdb to make presence/absence across kingdom. Remember to have strict cut-offs to filter out noise.

Thank you for your reply, Arun.

Can you please explain a bit? What is an NR database? (and where to find it?)

- In understand this is a string code, isn't it? -outfmt '6 qseqid sseqid evalue pident qcovs sgi sacc staxid ssciname scomname stitle'

Where to look about the usage of taxdb?

Sorry for so many questions, I am not familiar with the field.

Thank you again for your help

BLAST will work. NR is the non-redundant data set so if there are 10,000 identical sequences from 10,000 human samples in the database for example, only one is used.

For what you describe, you will want to check each kingdom separately. See attached screen shot. You can ask to search for the protein just in green plants for example. If you search in all of the NR database with a mammalian protein, you will get more than 10,000 "top hits" in the vertebrates and you will not see the less similar matches in the plants. If you don't find it in the plants you might then tryp to search in vertebrates but exclude tetrapods. If not there, then search in tetrapods but exclude mammals, etc...

My query above:

>EFG

ELHLDIIVDRMRREFNVEANVGKPQVAYRETIRKTDVKSDYKHAKQSGGKGQYGHVVIELSPMTEEDRKNDNVKDDFLFINDITGGIIPKEFIPSVEKGLRETITSGPIAGFPVVGVKVKLVFGSYHDVDSSEMAFKLAASMAFKQGFAKASPVLLEPIMKVEIVSPEDYLGDVMGDVSRRRGVLQGQDDSPSGKII

is a bacterial protein, also found in plants and animals. In eukaryotes it is found in two versions, one that works in the chloroplasts and mitochodria (although it is encoded in the nucleus and imported to the organelles) and one that work with the cytoplasmic ribosomes.

>EF2 MVNFTVDQIRAIMDKKANIRNMSVIAHVDHGKSTLTDSLVCKAGIIASARAGETRFTDTRKDEQERCITI KSTAISLFYELSENDLNFIKQSKDGSGFLINLIDSPGHVDFSSEVTAALRVTDGALVVVDCVSGVCVQTE TVLRQAIAERIKPVLMMNKMDRALLELQLEPEELYQTFQRIVENVNVIISTYGEGESGPMGNIMIDPVLG TVGFGSGLHGWAFTLKQFAEMYVAKFAAKGEGQLGAAERAKKVEDMMKKLWGDRYFDPANGKFSKSANSP DGKKLPRTFCQLILDPIFKVFDAIMNFRKEETAKLIEKLDIKLDSEDKDKEGKPLLKAVMRRWLPAGDAL LQMITIHLPSPVTAQKYRCELLYEGPPDDEAAMGIKSCDPKGPLMMYISKMVPTSDKGRFYAFGRVFSGV VSTGLKVRIMGPNYTPGKKEDLYLKPIQRTILMMGRYVEPIEDVPCGNIVGLVGVDQFLVKTGTITTFEH AHNMRVMKFSVSPVVRVAVEAKNPADLPKLVEGLKRLAKSDPMVQCIIEESGEHIIAGAGELHLEICLKD LEEDHACIPIKKSDPVVSYRETVSEESNVLCLSKSPNKHNRLYMKARPFPDGLAEDIDKGEVSARQELKA RARYLAEKYEWDVAEARKIWCFGPDGTGPNILTDITKGVQYLNEIKDSVVAGFQWATKEGALCEENMRGV RFDVHDVTLHADAIHRGGGQIIPTARRCLYASVLTAQPRLMEPIYLVEIQCPEQVVGGIYGVLNRKRGHV FEESQVAGTPMFVVKAYLPVNESFGFTADLRSNTGGQAFPQCVFDHWQILPGDPFDNSSRPSQVVAETRK RKGLKEGIPALDNFLDKL

It is useful to do some searches with proteins like this, to see what level of similarity might be expected between plants and animals. EFG is somewhat similar to EF2, but EF2 from plants is highly similar to EF2 from mammals.

Thank you for your replies, guys!

Arun Prasanna I am correct you are talking about the command line BLAST version (for sequences downloaded to your own server?). I could not find in NCBI where I could insert the options string.. But thank you for your help!

Brian Thomas Foley thank you for the visual explanation. The shortest intro to Bioinformatics;) I am right I have to manually pick all taxa I would like to search in? Can you tell me please about the results visualization options? Is it possible to have a sort of "tree" in which the presence of my protein in all the analyzed taxons can be visually assessed? Thank you!

Vad Pérez , it is not easy to guess how to help you when we have no idea what type of protein you are interested in. There are many proteins required for all cellular life, such as ribosomal proteins, DNA and RNA polymerases, energy metabolism proteins, etc... Many of these proteins are very highly conserved over time, but others are quite diverse even though they shared a common ancestor and still retain the same function in all cellular life. Many other proteins shared a common ancestor but no longer have exactly the same function. So it is not always possible to draw a line between the presence or absence of a protein in a lineage of life.

For one example, vertebrates have a wide variety of "immune system" proteins. Many of these evolved from proteins that were used for other functions in invertebrates, such as self-nonself recognition. It is not easy to determine exactly where in the history of life a protein like that took on immunity roles in addition to or instead of just self-nonself roles.

Many people think of the Eukaryotes as being "Plants, Animals, Fungi and Protists", but it is a bit more complex than this because the Protists include a HUGE variety of life forms or lineages in comparison the the other 3 groups, each of which is essentially just one fancy lineage of Protist that became multicellular. If we find something that is shared between plants and animals it is probably also shared with fungi and most protists, or perhaps it was lost by the fungi and/or one or more lineages if Protist. For one example, proteins associated with meiosis and diploid/haploid sex are shared by almost all eukaryotes.

Brian Thomas Foley, thank you again for the explanation. I am interested on a protein of the so called claster of differentiation (CD) group.

I understand your point about the groups diversity. Actually, I expect some differential distribution of the protein among groups, and just wanted to test my guessing.

Will follow your suggestions and tell the results. Thank you again!

Maybe this paper can help you Chakraborty, C., George Priya Doss, C., Sharma, R., Sahana, S., & Nair, T. S. (2013). Does Computational Biology Help us to Understand the Molecular Phylogenetics and Evolution of Cluster of Differentiation (CD) Proteins? The Protein Journal, 32(2), 143–154. doi:10.1007/s10930-013-9466-5.

my working pipeline may be:

1) Fetch human CDx (from Swiss-Prot)

2) BlastP against NR

3) Download Taxonomy reports from Blast results

I am not sure if you will find this surface receptors in many kingdoms... in fact I hope only find them in animals

Thank you, Manuel Mendoza! Looks quite interesting!

The CD proteins, such as CD4, CD8 etc. are well known in mammalian immunology as cell surface markers on T-helper cells, Killer T cells, Macrophages and so on. They are differentially regulated, along with other cell surface marker molecules during maturation of these cells. All vertebrates have immune systems, but even within the mammals there are a lot of differences in immune systems.

The paper cited by Manuel Mendoza above seems to imply that the various CD proteins (CD3, CD4, CD22, CDn...) shared a common ancestor that was involved in immune cell function, but this is not at all clear. There are thousands of other types of cell surface proteins with cell membrane-spanning hydrophobic regions rich in hydrophobic amino acids such as Leucine and Isoleucine.

Looking at immune system function across the kingdoms, you are out of luck. Although plants and fungi do have mechanisms of dealing with pathogens such as bacteria and viruses, they do not have B-cell and T-cells, so they can't have maturation functions of B-cell and T-cells in the thymus the way mammals do. It is entirely possible that other kingdoms have cell surface molecules with membrane-spanning regions that have some similarity to the CD proteins of mammals, but whether or not they truly shared a common ancestor would be essentially impossible to prove.

For a few proteins that are essential for life, such as the ribosomal proteins, DNA polymerase, and so on (I am attaching a sample of ribosome-associated Elongation Factor Tu protein) there is a very high degree of similarity across all eukaryotes. But most other proteins are not so conserved in sequence, and it can be nearly impossible to "see" their shared ancestry. My opinion, looking at the CD protein alignment, is that we cannot be sure those proteins shared any common ancestor. Surely CD22 of mammals shared one ancestor (the lineage that lead to marsupials and mammals probably had a CD22), but whether or no CD4 shared and ancestor with CD22 way back in the days of worms and clams, before the vertebrates evolved from the worms, is a different question.

BLAST and similar tools cannot answer these questions, for a number of reasons. There is a difference between asking what is the ancestor of a protein or group of proteins (membrane-spanning cell surface proteins for example) and asking when did one or more of these proteins take on some function in the immune system.

An introduction to earthworm immunity can be found in this online book:

https://www.ncbi.nlm.nih.gov/books/NBK45034/

Brian Thomas Foley thank you again for clearing all points!

Well, I am not interested in all CD proteins as a group, but just in one. There are homologs of this protein in animals, but I was interested on plants.

I understand the function could have evolved. Fortunately, I have an homolog from a more primitive animal, so I want to use this last as a template for a BLAST search in plants (actually, I performed the Delta BLAST and found some similarities with putative fungal proteins with no even remote similarities detected in plants).

I use BLAST just to see if the "gene" is somehow conserved (no matter if its function has changed). Apparently there is no such "gene" in plants. Next, for guessing what happened with the function, I will be looking in plants for the domain I am interested in. Perhaps plants developed its own protein with the activity I am looking for (I think ― they didn't).

May I ask you some other related questions after performing some preliminary tests?

Thank you again for your time!

Brian Thomas Foley may I can contact you regarding another project (protein engineering)? Is email ok for you?

Yes, email is fine. [email protected]

Thank you, Brian Thomas Foley let me accomodate the info to send it to you.

Back to my question — just suggested by RGArticle Automated selection of homologs to track the evolutionary hi...

This is also very good ; it also shows species in which the protein homologs are absent Article OrthoInspector 3.0: open portal for comparative genomics

And about my question on the specific protein, the analysis has been already done previously..

Brian Thomas Foley, the promised result. Deleted research item The research item mentioned here has been deleted

A figure is pending..