Guided Tour - Molecular Genetics - Fall 2008 | BSCI 410, Study notes of Molecular biology

Download Guided Tour - Molecular Genetics - Fall 2008 | BSCI 410 and more Study notes Molecular biology in PDF only on Docsity! Molecular Genetics last update Sept. 17, 2008 Web resources -- Tour. page 1 of 6 ------------------------------------------------------------------------------------------------------------- This is a guided tour. The homework is separate. In fact, this exercise is used for multiple classes. The tour will take a few hours. Do it with a friend (or two or three)! The NCBI site can be slow if it's very busy; this is hard to predict. If you are using this exercise for a class, be sure to start well before the homework is due! The details of how the NCBI site is configured are constantlly changing. If specific steps in this guide don't work as expected, keep going, or try to recreate the result using the site as revised (e.g. if a button is not where I say it should be, look around). The following are exercises to be carried out using a web browser (e.g. Firefox, Netscape, Safari or Microsoft Internet Explorer). GUIDED TOUR -------------------------------------------------- NCBI DATABASES Go to http://www.ncbi.nlm.nih.gov. This is the National Center for Biotechnology web site. (To get to this URL: Depending on your browser and operating system you can open a new location with control- L, control-O, command-L or command-O, and then type in the address. You can also just enter the address on the top line, also known as the address bar). You can also get to this address, and other useful bioinformatics addresses, from my bioinformatics links page, http://www.clfs.umd.edu/labs/mount/Bioinformatics.html. Once at the NCBI site, notice the top line ("quicklinks bar"), which lists PubMed, All Databases, Blast, OMIM, Books, TaxBrowser and Structure. We will consider most of these in turn. First, click on All Databases. This page presents all of the NCBI databases on a single page and allows you to search across databases. Enter "hemoglobin" into the search window and click on "Go." You should see numbers next to icons corresponding to each of the databases. At the very top are Pubmed, PubMed Central, Books, OMIM and site search. By clicking on one of these links you can see information about hemoglobin in PubMed, in books, on OMIM, etc.. More specialized databases are listed below. The following exercises will introduce you to these databases. PubMed (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed). This is a publicly available database of literature related to medicine, which includes all of molecular biology and genetics. To see how it works, navigate to that site using the top toolbar (click on PubMed) and type in Dreyfuss Xing Matunis You should get a single article in Nucleic Acids Research on hnRNP F. Note the box next to "Display" (near the top) with "Abstract Plus" written in it. This is a pull-down menu. If you put your mouse over the one labeled " Abstract Plus " and hold the mouse button down, you will see that there are other choices: Summary, Brief, Abstract, Citation, etc.. Most of these are different formats for presenting the citation. I encourage you to try them out. Because Nucleic Acids Research participates in PubMed Central, the full text of this article can be accessed directly by clicking on the link above the citation (on the right side). If the journal were not freely available to the public, it would still be possible to click a link here and connect to the article. In general, if the University of Maryland subscribes you should be able to connect to the journal from any campus computer. Molecular Genetics last update Sept. 17, 2008 Web resources -- Tour. page 2 of 6 ------------------------------------------------------------------------------------------------------------- From off-campus, you would need to access the journal article through Research Port (http://researchport.umd.edu/) and log in with your ID (use the 14 digit number on the back). Try "Links" and then select "nucleotide (refseq)". On the next page click the number next to, "core nucleotide." You should now see: You can get to the GenBank entries for these sequences directly by clicking on the accession numbers. You can also get to these same entries via the Nucleotide database. Let's see how that works. On the pull-down menu next to "Search," you will see the selections PubMed, Protein, Nucleotide, Structure, etc.. Select "CoreNucleotide," and then type in NM_004966 and click "Go." This takes you to the same entry you got through MedLine. Now, click on NM_004966, and you will see the GenBank entry for the locus HUMHFP, which encodes the hnRNP F protein. ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nuccore&id=148470398) Please note how the sequence is provided with a great deal of associated information. Note the pull-down menu labeled "Display" with "GenBank" written selected. Using your mouse, you can see that there are other choices: GenBank(Full), FASTA, ASN.1, Summary, etc.. Select FASTA (on the "Display" pull-down menu). You will see the sequence alone in a standard format known as "fasta format." Molecular Genetics last update Sept. 17, 2008 Web resources -- Tour. page 5 of 6 ------------------------------------------------------------------------------------------------------------- Examine your results. You will first see a figure illustrating where each "hit" matches your query sequence. Below that, you will see something like this: You can see the alignment for each "hit" by clicking on the score. If you click on the bit score of the accession ref|067485 you will see an alignment of the query (NP_004957, human hnRNP F) with the hit (NP_067485, mouse hnRNP H). "Query" refers to the sequence you searched with, while "Sbjct" (subject) refers to the matching database entry. The top scoring "hit" is a perfect match of the query to itself. The G on the right of this list takes you to the "Entrez Gene" page for this protein (Gene ID 3185 is human hnRNP F). A new feature that you might want to examine is a distance tree of results (just above the list of hits). Give it a try! Look at the alignments!!!! Go back to the BLAST results (if necessary, use your Request ID) and return to the alignments. If you cruise through the list of hits, you will see that most of the top scoring hits are similar proteins in more closely related species. In order to evaluate the meaning of a match, you should pay close attention to both the percent identity and E value. Molecular Genetics last update Sept. 17, 2008 Web resources -- Tour. page 6 of 6 ------------------------------------------------------------------------------------------------------------- For example, find gi|25153681|ref|NP_741422.1|| and go to the alignment (by clicking on the score). You will see: The E value of 1e-164 means that you would expect 7 x 10-164 hits this good or better if you searched a database of random sequences the size of the database that you actually searched. Better matches yield lower E values. Generally, E values less than 0.05 have statistical significance, but the biological significance may be meaningless unless E values are below 1e-6 or even lower. Matches with scores this good are not rare when related proteins are compared. Many of the top hits in this search are extremely similar. You may want to find less similar proteins, in another species, for example. There are two ways to do this. First, go to the top of the blast output window and click on "Taxonomy Reports" This will generate a searchable window that presents the hits organized by the taxonomy of the species involved. As an alternative, you can limit your search in the first place. To do this, go back to the main BLAST window, and repeat your search. This time, you want to restrict your search to the fruit fly Drosophila melanogaster. Repeat the search after selecting Drosophila melanogaster under "Options" You may notice that the search takes much less time this time around! That is because you are searching a smaller database. Look at your output, including the alignment, and you will see a set of related proteins. How many do you think are likely to have a similar function? Many of the hits obtained with this search are proteins containing the RRM domain. You can see the location of RRM domains in hnRNP F by returning to the blast page and looking at the Conserved Domains hits, located just above the Request ID. Alternatively, return to the main NCBI page, http://www.ncbi.nlm.nih.gov/ and search Proteins using the accession NP_004957. Once you have the protein listing, you can click on "Conserved Domains," which is between BLink and Links on the upper right. You should see three RRM domains aligned with the hnRNP F sequence. If you click on any one of these, you will see an aligment of several specific RRM proteins with the query hnRNP F. You may want to explore links from this page. Note particularly the Pfam page for this domain (pfam00076). In this exercise, I have taken you through just a few of the available programs in great detail. Play around. Explore the site yourself. You can also get help online by clicking on links within the BLAST pages. There are also help and tutorial links on the NCBI page themselves. There are other sites that this tutorial did not cover, like Gene Ontology (www.geneontology.org/), that you may also want to look at; many of these are listed on my links page. Good luck and have fun!