intro

Date: April 20^th, 2005

Title: Using Bioinformatics Research Tools to Increase Our Knowledge About Genes and Proteins for Cell and Molecular Biology

Objective: The purpose of this Bioinformatics project was to familiarize ourselves with Bioinformatics tools to research a specific Gene and Protein. I was provided with a partial amino acid sequence taken from Hela cells as follows:

GVAGAHGLLCLLSDHVDKRILDAAGANLKVISTMSVGIDHLALDEIKKRGIRVGYTPDVLTDTTAELAVSLL
LTTCRRLPEAIEEVKNGGWTSWKPLWLCGYGLTQSTVGIIGLGRIGQAIARRLKPFGVQRFLYTGRQPRPEE
AAEFQAEFVSTPELAAQSDFIVVACSLTPATEGLCNKDFFQKMKETAVFINISRGDVVNQDDLYQALASGKI
AAAGLDVTSPEPLPTNHPLLTLKNCVILPHIGSATHRTRNTMSLLAANNLLAGLRGEPMPSELK

With this data, my objective was to discover the rest of the sequence, and consequently which Gene and Protein it codes for. I was able to use tools such as NCBI and BLAST among others on the internet to find its identity.

Background: Bioinformatics is the "marriage" of Molecular Biology and Computer Science. Using proper tools, scientists can collect and organize entire genomes. The human genome alone has three billion pieces of information, so there is a need to manage tremendous amounts of data. Through Bioinformatics, the relationships between genes and proteins can be studied in depth. These relationships can give rise to gene discoveries, diagnosis and treatment of diseases, and creation of new and custom drugs¹.

In order to solve a partial amino acid sequence, all the information at hand must be taken into account. (For example, in this case I was told that this amino acid sequence was taken from Hela cells. Hela Cells² are derived from humans. This comes in handy later when trying to find the best match for the sequence. If one match is human and the other is non-human, then we know the human gene is the correct one.) When the partial sequence is entered, the database searches for other genes or proteins with a similar sequence. The database also uses an algorithm to formulate an "e-value" (error-value) which says how many sequences from the database when aligned with yours will produce the same "score" randomly. The highest (negative) e-value is the most closely related and thus the best choice. When the correct gene is found, links from that page can provide information about the protein it codes for and its purpose, etc.

Tools and Methods: Before starting this project, I had to take a step back to gather the information at hand and the tasks I would need to perform. Initially, I knew I had an amino acid sequence, which would require a protein-protein search with BLAST. Secondly, I knew that the protein and gene that code it were from a human. Below, I have outlined the methods I used and have included the actual results where possible.

1) I performed a protein - protein search with BLAST from the NCBI website. From the results, I was able to see possible conserved domains. The next step was to click on the 'Format' button, the results are the related proteins to the sequence I entered into BLAST. The results with the highest negative 'e-value' is most probably the protein I was looking for. By Clicking on The blue 'G' button it takes me to the Gene, GRHPR, which codes for that protein. From the gene page all the information necessary is accessible by the click of a button.

2) I needed to find the sequence of the gene in order to design a map of it. By scrolling down the gene page, I found the related genomic sequence to be AF146689³. From here I was able to map out the Exons, Introns, CDS and UTR. Links are also available to PUBMED and OMIM where health and medical information can be found pertaining to my gene. A link was also available for the mRNA sequence for this gene.

3) Back to the original gene page, scrolling down again I found the protein whose amino acid sequence I was given, AAD46517⁴.

4) The last step was to find a structure for the protein. By clicking on 'BLink' I was able to see a list of proteins related to mine. By click '3-D structures' on the top I was given a list of proteins whose 3-D structure are known, unfortunately mine was not on that list. No sweat though, the protein listed at the top is the closest match to mine that has a known 3-D structure. By clicking on the blue dot it brought me to that proteins page and showed me a comparison of the two proteins. Next, by clicking the 'MMDB: 1GDH_B' button it brought me to another screen from which I clicked 'PDB: 1GDH' which brought me to the protein databank from which by clicking on view structure I was able to obtain a picture of the proteins structure.