Protein Analysis Subcellular Localization Prediction

Protein (or in general, proteome) Analysis Subcellular Localization Prediction is a process (usually through the use of web-based software) of predicting the location or destination of a protein within the cell using only the protein sequence as its inputs.

In the central dogma of molecular biology, the major copy of the message for protein synthesis is contained in the DNA. The message is transcribed into mRNA and then transported to the cytosol where translation of the copies of mRNA into proteins happens. Several post-translational modification processes happen before the cell could actually use the modified protein. This protein is then transported to a particular organelle (intracellular) or outside the cell (extracellular) where the protein is needed. Proteins are then likened to letters with proper address and stamps to deliver it ont he proper destination. Likewise, proteins should have proper address to ensure its delivery to the proper localization. The destination of various protein sequences is predicted by these subcellular localization prediction servers.

Protein Localization Prediction is important to predict the possible location or destination of a particular protein. Take note, however, that this is still a prediction. An actual wet-lab procedure is preferred to do actual analysis of the protein's location. Since the prediction servers present possible location of the protein, then it also gives us insights of the possible function of the protein sequence.

One web-based program is the PA Specialized Subcellular Localization Server v2.5 found at Proteome Analyst Research Group, University of Alberta. Several related servers are also available for protein subcellular localization prediction. A list can be found at PSORT.

Additional recommended knowledge

Daily Sensitivity Test

What is the Sensitivity of my Balance?

What is the Correct Way to Check Repeatability in Balances?

Sample run of the tool

For our particular purpose, we will use the PA-Subcellular Localization Prediction server found at the University of Alberta. To do a sample run of the tool, let us consider the Oxalate oxidase (OxO) enzyme from rice (NCBI). To give you an introduction, OxO are involved in plant defence to pathogen infection such as powdery mildew of barley and wheat. It catalyzes the following reaction:

Oxalate + O2 + 2H+ = 2 CO2 + H2O2. 

The hydrogen peroxide, H₂O₂, produced may play an important role in several aspects of defense mechanisms. In rice, this enzyme has been mapped at chromosome 3.

1. enter an Analysis ID, like, "OxO".
2. Now, copy the FastA sequence below and paste it on the PENCE PA-SUB server opposite the "FastA Sequences" box.

>gi|108710532|gb|ABF98327.1| Oxalate oxidase 2 precursor, putative, expressed [Oryza sativa (japonica cultivar-group)]
MEHSFKTIAAGVVIVVLLLQQAPVLIRATDADPLQDFCVADLDSKVTVNGHACKPASAAGDEFLFSSKIA
TGGDVNANPNGSNVTELDVAEWPGVNTLGVSMNRVDFAPGGTNPPHVHPRATEVGIVLRGELLVGIIGTL
DTGNRYYSKVVRAGETFVIPRGLMHFQFNVGKTEATMVVSFNSQNPGIVFVPLTLFGSNPPIPTPVLVKA
LRVDAGVVELLKSKFTGGY

3. Click the "Plant" radio button on the "Organism Type" table. Click the "Begin Analysis" button. An intermediate page will show up. You may view the file by clicking the appropriate hyperlink or you may download it in *.csv file.

Result

A screenshot of the result of our sample run is shown at the right pane. It contains statistics on the protein we submitted, along with BLAST results and a Subcellular Prediction.

The positive prediction gives as the most probable location of the protein sequence, in this cae, the protein OxO is secreted extracellularly (100%), probably on the cell wall, or transported to vacuole (90.346%).

For references on interpreting the results, click any of the following [1] [2] [3]

Try this

Perform the same procedure as above but use the following sequence as an entry. This amino acid sequence is Ribulose bisphosphate carboxylase (RUBISCO) large unit. From this information, we can already predict its possible location.

>gi|476752|gb|AAB02583.1| rubisco large subunit
MSPQTETKASVGFKAGVKDYKLTYYTPEYETKDTDILAAFRVTPQPGVPPEEAGAAVAAESSTGTWTTVW
TDGLTSLDRYKGRCYHIEPVVGEDNQYIAYVAYPLDLFEEGSVTNMFTSIVGNVFGFKALRALRLEDLRI
PPTYSKTFQGPPHGIQVERDKLNKYGRPLLGCTIKPKLGLSAKNYGRACYECLAGGLDFTKDDENVNSQP
FMRWRGRFVFCAEAIYKSQAETGEIKGHYLNATAGTCEEMIKRAVFARELGVPIVMHDYLTGGFTANTSL
AHYCRDNGLLLHIHRAMHAVIDRQKNHGMHFRVLAKALRMSGGDHIHAGTVVGKLEGIFFFFFFFFFFFL
RKELYVCVIIQNKHVVQGVAVHACRISSSLWYTYPSKRSKTYETNTQEEES

Other Subcellular Localization Servers

• PSORT
• SubLoc.
• TMHMM

References

• PSORT_Wikipedia
• PSORT
• SubLoc.
• TMHMM

If you include this web-based tool in any of your research, cite:
Z. Lu, D. Szafron, R. Greiner, P. Lu, D.S. Wishart, B. Poulin, J. Anvik, C. Macdonell, and R. Eisner, Predicting Subcellular Localization of Proteins using Machine-Learned Classifiers, Bioinformatics 2004 20(4):547-556.