Exercises in Data Retrieval and Using Blast Searches
[The NCBI web site and its parts are updated periodically, therefore results given below may change with time.]
Step by step instructions with screen shots are given at the end of each exercise.
Instuctions not in the step by step section of the handout are given in green type.

1.	What is the breath of information available at NCBI on cystic fibrosis in humans?
2.	Besides the cystic fibrosis transmembrane conductance regulator gene (CFTR), what other genes are associated with cystic fibrosis in humans and what are their roles in the disease?
3.	Nocturnal asthma is associated with what gene in humans? What are the RefSeq codes for this gene's mRNA and protein sequences? On the GenBank accession pages, data can be displayed in different formats. What is the difference between default and FASTA formats for these sequence files? How can these RefSeq codes be used to search for similar sequences in other species? What are the results of such a search?
4.	Are there any solved protein crystal structure(s) for the nocturnal asthma gene. Does the structure include the transmembrane segments? Are the found structure protein and the nocturnal asthma protein closely enough related to believe the results?
5.	Find proteins that are known to contribute to pulmonary artery hypertension and determine if animal models exist in which the disease can be studied. Can a full length dog protein sequence be found?
6.	How conserved are the ATP2A2 proteins across vertebrate species? Should all the available protein sequences be used to make this assessment?
7.	Are there knockout mice available to study the AGPAT6 gene? How would you order one of these cell lines?

#1 What is the breath of information available at NCBI on cystic fibrosis in humans?

Hints: Do a All Databases search at NCBI (http://www.ncbi.nlm.nih.gov). Then repeat the search narrowing the returned hits to human.

Answer: Lots of data to be explored.

#1 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov,

enter cystic fibrosis in the for box

and click Go.

2. The returned Entrez page is organized with literature matches in the top box, sequence based information in the middle one and NLM's resources in the bottom box. Items of possible interest are denoted by numbers in a white box next to a topic title.

There are 146 OMIM entries [catalog of human genes and genetic disorders].

OMIM background information with links off to help and frequently asked questions

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=OMIM

The sequence based information section contains data on cystic fibrosis from all species.

Some of these topics are sensitive to the addition of species information to the search query. OMIM and MeSH are examples of this.

3. To restrict the sequence data to that from humans, add homo sapiens to the current cystic fibrosis in the Search across databases box and click Go.

4. The number of matches for nucleotide, protein, gene topics have decreased, but, there still is a large number of items to sift through.

#2 Besides the cystic fibrosis transmembrane conductance regulator gene (CFTR), what other genes are associated with cystic fibrosis in humans and what are their roles in the disease?

Hints: Perform an Entrez Gene search (http://www.ncbi.nlm.nih.gov) to find the other genes and their function or relationship to the disease.

Possible answers:

S100A8	-	cystic fibrosis antigen
TGFB1	-	mutations modify severity of pulmonary disease in cystic fibrosis patients
TGFB1	-	protein expression correlates with portal tracts showing histological abnormalities associated with cystic fibrosis liver disease
GOPC	-	CFTR binding
ADRB2	-	2002 polymorphisms contribute to clinical severity and disease progressionin cystic fibrosis 2005 - transfected beta3 not beta2-adrenergic receptors regulates CFTR activity via new pathway
ABCB1	-	study to see how the common cystic fibrosis mutation might disturb transmembrane segments of the protein using ABCB1 as a model ABCB1 expression increases ATP release in respiratory cystic fibrosis cells potential clinical benefits discussed

#2 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov,

change the Search option from All Databases to Gene using the pull down menu,

enter homo sapiens cystic fibrosis in the for box

and click Go.

Click on at least five diverse hits below the CFTR gene, finding out their relationship to cystic fibrosis. Ignore any gene that doesn't have a NCBI Reference Sequences (RefSeq) section.

Here is what one of the results pages looks like.

Click on the PUBMED links in the middle of the page and scan through the titles for mention of cystic fibrosis.

The given example (S100A8) had 3 pages of references and the cystic fibrosis ones were on the last page.

If no papers are listed with cystic fibrosis in the description, check out the MIM link at the top of the page or the links in the GeneOntology section.

Possible answers:

CFM1	-	no RefSeq data (ignored)
CFM2	-	no RefSeq data (ignored)
S100A8	-	cystic fibrosis antigen
TGFB1	-	mutations modify severity of pulmonary disease in cystic fibrosis patients
TGFB1	-	protein expression correlates with portal tracts showing histological abnormalities associated with cystic fibrosis liver disease
GOPC	-	CFTR binding
ADRB2	-	2002 polymorphisms contribute to clinical severity and disease progressionin cystic fibrosis 2005 - transfected beta3 not beta2-adrenergic receptors regulates CFTR activity via new pathway
ABCB1	-	study to see how the common cystic fibrosis mutation might disturb transmembrane segments of the protein using ABCB1 as a model ABCB1 expression increases ATP release in respiratory cystic fibrosis cells potential clinical benefits discussed

Note there is another database that is relevant for getting clinical information, Online Mendelian Inheritance in Man (OMIM or MIM). Althought OMIM can not be searched via an actual sequence, it does allow searching by gene symbol, chromosome location, keywords or other features.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=omim

#3 Nocturnal asthma associated with what gene in humans? What are the RefSeq codes for this gene's mRNA and protein sequences? On the GenBank accession pages, data can be displayed in different formats. What is the difference between default and FASTA formats for these sequence files? How can these RefSeq codes be used to search for similar sequences in other species? What are the results of such a search?

Hints: Do a Gene search at NCBI (http://www.ncbi.nlm.nih.gov), record the codes. Compare the formats of the mRNA and protein sequences. Run a BLAST search, (http://www.ncbi.nlm.nih.gov/BLAST/).

Answers:

gene - ADBR2 adrenergic, beta-2-, receptor, surface

Product (protein) NP_000015

FASTA format is very concise, limited to the actual sequence and an identification line that starts with a > symbol. The default format is very verbose, giving all sorts of reference details about the sequence and a version of the sequence that is more easily read by the user.

BLAST searching allows for different types of data entry including the use of accession codes (such as a RefSeq accession code).

ADBR2 contains the 7tm_1 conserved domain signature which is highly conserved across species.

#3 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov,

change the All Databases Search option to Gene using the pull down menu,

enter homo sapiens nocturnal asthma in the for box

and click Go.

2. Check the resulting hits to insure that the summary information on the gene mentions that various types of changes in this gene are associated with the disease.

Here are the summary sections of the top three hits.

Only the first one contains a reference to nocturnal asthma.

ADBR2 adrenergic, beta-2-, receptor, surface

3. Scroll down the page to the NCBI Reference Sequences (RefSeq) section. Record the mRNA sequence and Product (protein) codes:

The required information:

NM_000024

NP_000015

4. Click on the mRNA code to see the data on the actual mRNA sequence data. Scroll down the page taking in the format of the information presented.

mRNA sequence link

5. Scroll back to the top of the page and change the Display option from GenBank to FASTA.
The format automatically changes. Note the difference. FASTA format is the sequence format required by many database searching programs.

background information on fasta format

6. Click back to the Entrez Gene page and repeat this process with the protein code.

protein link

7. After noting the difference, click on the NCBI logo at the top of the page.

From the blue navigation bar on the main NCBI page,

click on BLAST.

8. From the main BLAST page, click on protein blastp in the Basic BLAST section.

In the blastp suite page, click on the ? icon above the large box to find out about what sort of inputs this form accepts. Clicking the more... link provides additional details.

After reading the presented information, close the popup window and the original "?" block, then enter the protein code into the Enter Query Sequence box. Once this is done, information appears in the Job Title box.

In the Choose Search Set section of the page, change the Organism button to Custom and start to enter the term Vertebrata into the Organism box.

As the term is entered, matching Entrez terms start to appear. When enough of the word is entered to find the desired term, select this term from the list.

Clicking the BLAST button at the bottom of the page starts the search.

If results are to be displayed in a new window, click on the "Show results in a new window" box prior to clicking the BLAST button.

Protein searches gets around the problem of multiple codons coding for the same amino acid that impacts nucleotide searches. However, depending in the information sought, this is not always possible.

9. It may take a few seconds for the search to be completed. While waiting, click on the 7tm_1 in the image to find out about the conserved domain that was found in the sequence.

Initial Conserved Domains page.

Clicking on the + symbol at the beginning of the green highlighted line produces the full version of the page.

7tm_1 indicates that the protein being search with contains the transmembrane receptor signature of the rhodopsin family of transmembrane proteins.

This signature is located from residue 50 to 325 in the sequence.

Close the popup window.

Note the length of the query sequence, this may be given on the Query line. 413 letters

Wait for the results page to appear.

Here is a screen shot of the top part of the Blast results page.

10. Scroll down the results page past the image with its colored horizontal bars to the Sequences producing significant alignments section.

Scores are based on the length of the query sequence and the size of the database. Short sequences will never produce great scores. To get a E value of 0.0 requires a match of at least 330 characters. A very long sequence could easily have a match this long and still not have a match that covers a significant portion of the query sequence. Always look at the resulting alignment. The mathematics of the process can sometimes result in the strange ordering of hits.

A hit line gives the GI number of the match sequence, the database it is from, the accession code used for the sequence, the description of the sequence from the database, its Bits score and finally the E value. Hits in the list are ordered their E value, then their Bits score which reflects the length of their actual match. Enough of the description may be given to see what species the sequence is from.

Clicking on the link given on the left side of a hit line goes off to the actual sequence information.
Clicking on the right side link moves down to the alignment data for that hit.

Notice that there are over 30 hits with an E value of 0.0 at the top of this list and that the protein code entered is not at the top of the list. There are about 130 hits in the list which mention ADBR2, beta-2 adrenergic receptor or variations thereof before sequence description changes to something else. The first 12 hits are all from man with from 0 to 2 mismatches in the alignment.

NCBI used to make an effort to remove redundant sequences, but the size of the database increased to such an extent that it was no longer possible to do this quickly enough so that it wouldn't impact the processing of new data.

When an accession code begins with XP_, it means that the data is the results of an automated analysis process. This situation usually occurs when a genome sequencing project is first being analyzed. These sequences have not been checked for accuracy and can be much longer or shorter than their homologs from more mature genome studies. These sequences usually have their description start with PREDICTED:.

Check out some of the hits beyond the 0.0 E values and determine where the match is actually taking place within the query sequence.

PREDICTED: similar to beta-2 adrenergic receptor, [Gallus gallus] Length=397

results alignment section
12 - 396
beta-2 adrenergic receptor [Homo sapiens] Length=275

results alignment section
52- 326
beta-2 adrenergic receptor [Macaca mulatta] Length=275

results alignment section
52 - 326
beta-2 adrenergic receptor [Hylobates concolor] Length=275

results alignment section
52 - 326
beta-2 adrenergic receptor [Ateles fusciceps] Length=275

results alignment section
52 - 326
....
beta-2 adrenergic receptor [Hippopotamus amphibius] Length=275

results alignment section
52 - 326

The match is happening in the 7tm_1 region of the sequence which appears to be highly conserved.

#4 Are there any solved protein crystal structure(s) for the nocturnal asthma gene. Does the structure include the transmembrane segments? Are the found structure protein and the nocturnal asthma protein closely enough related to believe the results?

Hints: Use the protein accession code from the previous exercise and run a protein BLAST search (http://www.ncbi.nlm.nih.gov/BLAST/). This time, instead of using the default database, use the swissprotein database and a structure database. Compare the two human proteins with BLAST 2 SEQUENCES ( http://www.ncbi.nlm.nih.gov/blast/bl2seq/wblast2.cgi) to make the decision.

Answers:

swissprotein - ADBR2_human

transmembrane segments:	1. 35 - 58	3. 107 - 129	5. 197 - 220	7. 306 - 329
	2. 72 - 95	4. 151 - 174	6. 275 - 298

pdb match - bovine Rhodopsin covers the transmembrane segments region

comparison - The results page shows that the two human proteins are not very related to one another. Solving transmembrane protein structures is very difficult. Perhaps a few more structures for this protein family should be obtained before believing the alignment results.

#4 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov, and from the blue navigation bar click on BLAST.

2. From the main BLAST page, click on protein blast in the Basic BLAST section.

In the blastp suite page, click on the ? icon at the end of the Database line in the Choose Search Set section to find out information about the databases that can be used in this protein BLAST search. Clicking on the more... link provides additional information. Once a suitable structure database name has been located, close the more... page and re-click on the "? icon" to close the information block.

From the list given the structural database to use is pdb. The swissprotein database was also listed.

Of the protein databases, swissprotein is considered to have the best annotation. One of the features they report is transmembrane segment locations when available or predicted.

3. Change the Choose Search Set Database option from nr to swissprotein using the pull down menu,

enter the previously found RefSeq protein accession code into the Enter Query Sequence box.

To speed things up and reduce the size of the output file, restrict the organism searched to humans by changing the Any organisms option to Human by clicking the Human button in the Organism line.

Start the run by clicking the BLAST button.

4. At the top of the actual results page, click on the "Reformat these Results" link.

This leads off to a form which allows the changing of the produced results.

The number of descriptions, lines in the image and alignments can be restricted using the Descriptions:, Graphical overview: and Alignments: pull down menus. Restrict these three options to 10 each and then click on the View report button near the top of the page.

Here is the top part of the results page.

5. Scroll down the results to the significant alignments section and click on the sequence link containing the term ADRB2_HUMAN. It should be the first one on the list.

link to ADRB2_human

6. Scroll down the swissprotein data file to the FEATURES section. Then read through the listed features to find those regions called "Transmembrane region".

The first Transmembrane region from the data file

There are a total of 7 TMD regions

transmembrane segments: 1. 35 - 58 3. 107 - 129 5. 197 - 220 7. 306 - 329

2. 72 - 95 4. 151 - 174 6. 275 - 298

7. Return to the protein blast page, re-enter the RefSeq accession code if necessary, and change the database to be used to pdb,

return Organism to its default Any value,

and then click BLAST button.

8. Wait until results page appears.

9. The best hit comes from Bovine Rhodopsin, but it is not very strong.

The alignment does cover the entire area containing the transmembrane segments.

10. Return to the main NCBI page, change All databases to Gene and enter rhodopsin homo sapiens in the for box and click Go.

11. The gene symbol for rhodopsin is RHO from the result of this search.

Click on the RHO link to get to the Entrez Gene page. Scroll down the page to find the RefSeq protein accession.

The protein code is

Product NP_000530

12. Return to the main NCBI page,

click on BLAST from the blue navigation bar.

This time on the main BLAST page, click on the "Align two sequences using BLAST (bl2seq)" link in the Specialized BLAST section.

13. In the BLAST 2 SEQUENCES page,

change the Program from blastn to blastp, enter the protein accession code for ADRB2 in the sequence 1 box and the RHO code in the sequence 2 box and click Align.

The results page shows that the two human proteins are not highly related to one another.

Solving transmembrane protein structures is very difficult. Perhaps a few more structures for this protein family should be obtained before believing the alignment results.

#5 Find proteins that are known to contribute to pulmonary artery hypertension and determine if animal models exist in which the disease can be studied. Can a full length dog protein sequence be found?

Hints:

Determine the proteins associated with the disease by doing an Entrez protein search Choose the top hit and check its length. Use the sequence in a NCBI protein BLAST search to try to find possible model animals. Also use this sequence to find dog sequences. Compare any found dog sequences with the human sequence you started with. You can do this with a BLAST 2 SEQUENCES (http://www.ncbi.nlm.nih.gov/blast/bl2seq/wblast2.cgi) run.

Answers:

protein - bone morphogenetic protein receptor type II

lengths - human: 1038

rhesus monkey, mouse and rat would be possible animal models

best dog matches	XP_536035 759 aa
	XP_851509 248 aa

complete dog sequence - No, there appears to be about a 30 residue gap between the two dog sequences that needs to be filled before the sequence is complete.

#5 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov, change the All Databases option to Protein and enter the term pulmonary artery hypertension homo sapiens into the for box

and click Go.

The results of the search.

2. Choose the top hit, in this case NP_001195. Clicking on the link will take you to the page for the protein.

NP_001195 data file

3. Check out the length of the protein. The length is the second item on the "LOCUS" line.

the protein is 1038 residues long

(The sequence to be used in the search is 1038 residues long)

4. To obtain the sequence of NP_001195 for the BLAST search, change the Display option of the page from GenPept to FASTA using the pull down menu.

This automatically changes the format to FASTA.

Copy this data, starting with the ">" and continuing to the end of the sequence.

5. Click on the NCBI logo in the upper left-hand side of the screen.

6. From the main NCBI page click on BLAST in the blue navigation bar.

On the main BLAST page, click on protein blast in the Basic BLAST portion of the page.

7. Paste your sequence into the Enter Query Sequence box,

be sure that the Choose Search Set parameter are at their default values (database nr and organism Any)

and then click the BLAST button.

8. Wait until the results page appears.

9. Check out the best hits with a description that appears to be correct,

Checking out the best hits to determine the quality of the matches and the species results in the following information.

XP_001101663	[Macaca mulatta]	1038 residues	Identities = 1029/1038 (99%)
NP_031587	[Mus musculus]	1038 residues	Identities = 1003/1038 (96%)
XP_001065181	[Rattus norvegicus]	1038 residues	Identities = 998/1038 (96%)
NP_001001465	[Gallus gallus]	1031 residues	Identities = 918/1021 (89%)
XP_617592	[Bos taurus]	914 residues	Identities = 868/898 (96%)
AAB39883	[Xenopus laevis]	1048 residues	Identities = 784/1028 (76%)

These results would indicate that rhesus monkey [Macaca mulatta], mouse and rat would all be good animal models in which to study this gene and it function.

10. Return to the blastp suite submission page and change the Organism option in the Choose Search Set section from Any to Canis familiaris by clicking the Custom button and starting to enter this term into the field. Highlight the term when it appears on the list.

Click the BLAST button to start the run.

Confirm that the hits are from dog and determine how close the length is to that of the starting human sequence. The first two look the most likely.

11. Return to the main BLAST page,

and click on the "Align two sequences using BLAST (bl2seq)" in the Specialized BLAST section.

12. On the BLAST 2 SEQUENCES page, change the Program from blastn to blastp, enter the protein accession code for the human protein sequence in the sequence 1 box and the first dog code in the sequence 2 box,

click Align.

For XP_536035 the alignment goes from 1 - 758 of the human sequence and the identity is 97%.

Return to the BLAST 2 SEQUENCES page, change the code in sequence 2 box to that of the second hit

click Align.

For XP_851509 the alignment goes from 791 - 1038 and the identity is 93%.

There appears to be about a 30 residue gap in the two dog sequences that needs to be filled before the sequence is complete.

#6 How conserved are the ATP2A2 proteins across vertebrate species? Should all the available protein sequences be used to make this assessment?

Hints:

Determine the easily available protein sequences with a Gene search at NCBI (http://www.ncbi.nlm.nih.gov). Record the species found in the search. Follow this with a BLAST search (http://www.ncbi.nlm.nih.gov/BLAST/) with one of the sequences to tease out others in the swissprotein database. Repeat this search on the nr database to get more sequences. Do a CLUSTALW multiple alignment on this data at EBI (http://www.ebi.ac.uk/clustalw/) and make your assessment. Observe the quality of the alignment throughout its length.

The ATP2A2 proteins are very highly conserved across vertebrate species. However, any found partial sequences should not be included in the alignment.

#6 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov, change for Search option from All Databases to Gene, enter ATP2A2 in the for box

and click Go.

2. Work through the list of exact gene name matches.

First, go to the Entrez Gene page for each entry. Scroll down this page to the NCBI Reference Sequences (RefSeq) section. If this section is there, record the RefSeq name of the protein product and check to see how long the sequence is by clicking on the link. Record the length as well. If a Related Sequences is only available, record the protein's name and check on its length by clicking on the link.

There are 8 exact gene name matches:

frog [Xenopus laevis]	AAH77920	574
human [Homo sapiens]	NP_001672,NP_733765	997,1042
dog [Canis familiaris]	NP_001003214	997
mouse [Mus musculus]	NP_033852	998
chimp [Pan troglodytes]	three predicted XPs
rat [Rattus norvegicus]	NP_058986	1043
cat [Felis catus]	NP_001009216	997
chicken [Gallus gallus]	AAA49066	997

From the information contained in the human Entrez Gene page, there are two known isoforms for this gene. One whose length is 997 and another whose length is 1042 (NP_733765). Use the longer one in your analysis runs.

3. Use this RefSeq name to do a BLAST search on all vertebrate sequences, by returning to a page with the NCBI logo on it and clicking on that logo.

From the NCBI main page, click on BLAST in the blue navigation bar.

Click on the protein blast link in the Basic BLAST section to get to the protein search form.

Enter your RefSeq name into the Enter Query Sequence box, change Any to Vertebrata, in the Organism line and change the searched database to swissprotein,

and then click the BLAST button.

5. Wait until a results page appears.

6. Scroll down the page to the significant alignments section.

Check out the top hits to see what the swissprotein family code is for the gene symbol ATP2A2.

AT2A2_human [human]	1042
AT2A2_pig [pig]	1042
AT2A2_rat [rat]	1043
AT2A2_rabit [rabbit]	1042
AT2A2_mouse [mouse]	1044
AT2A2_chick [chicken]	1041
AT2A2_canfa [dog]	997
AT2A2_felca [cat]	997

	The swissprotein family name appears to be AT2A2.
7.	Repeat this process using the nr databases to find other possible sequences. Return to the Blast submission page and change the database from swissprotein to nr using the pull down menu.

Check out the results to find full sequences (around ~ 1038 residues) from more species

Possible additional sequences are:

XP_001141715 chimp [Pan troglodytes] 1042 Identities = 1037/1042 (99%)

CAJ42045 horse [Equus caballus] 1042 Identities = 1032/1042 (99%)

AAH98958 frog [Xenopus laevis] 1042 Identities = 961/1042 (92%)

From this list of possible sequences, those starting with XP_ probably won't work at the ClustalW site.

8. Go to the CLUSTALW site (http://www.ebi.ac.uk/clustalw/).

Data can be entered in a number of ways on this page. But, notice the new feature at the top of the page.

This feature allows the EBI local databases to be searched for sequences of interest. Replace the Enter Text Here in the box with a swissprotein accession code

and click Go.

Protein sequences are needed for the alignment and there is one listed on the page. Click on the Protein Sequences link.

There is one match in the UniProt database, to see the actual data file, click on the in UniProt format link in the View line.

What appears next is the local copy of the AT2A2_HUMAN data file. Copy the entire data file and then return to the starting EBI ClustalW page and paste it into the Enter or Paste box.

Repeat this process with all 8 of the found swissprotein family members. You will need to scroll down to the bottom of the Enter or Paste box each time to add your new data. Be sure that there is a cursor below the // of the last entry, below this you paste in your new data.

After all your sequences have been pasted in,

click the Run button.

9. The waiting page, that lets you known that the site is working on your alignment,

is replaced by your results.

In the first part of the page is a box listing the ClustalW Results with a JalView button. Click it and see if a colored image comes up. [This may not work with all browsers.]

10a. If the image appears, the columns are colored according to the residue type. A column that is all the same color and which contains the same character in each row is a conserved position. Care needs to be taken with the A,I,L,M,V locations, as the site considers these residues interchangeable, but they are not. Move across the alignment and see how it changes with position.

10b. If the image doesn't appear, move down the page to the Alignment portion.

Click the Show Colors button.

This produces a new page where the alignment position characters are colored according to residue type.

Again, care needs to be taken with the A,I,L,M,V locations. All the positions with an "*" in the line below the members of the alignment have the same character in that location.

11. Depending on how careful you where in selecting sequences, there could be possible changes.

There could be two types of proteins here based on length. Looking at the human Entrez Gene page showed that there were two isoforms of this gene. They only seem to differ at the C terminus end, where one form is longer than of other.

The end of the alignment looks like this.

If you have more than one sequence from the same species, you might want to drop the shorter one.

You might want to try to add additional sequences to this alignment using the codes that were collected. There is a frog code, AAH98958 you could search for and add to your alignment. EBI doesn't have predicted RefSeq data, but it might have most other protein sequence codes. You just have to try a search and see.

Adding the frog data changed the aligned data. The least conserved area is at the very end, between 993 and 1044. Only the cat and dog sequences seem to short. Perhaps their sequence data is not for the same isoform as the others.

The members of the alignment appear to be very highly conserved when the same isoform is used across species..

#7 Are there knockout mice available to study the AGPAT6 gene? How would you order one of these cell lines?

Hints:

Find the mRNA FASTA formatted sequence for the AGPAT6 mouse gene by doing an Entrez Gene search at NCBI (http://www.ncbi.nlm.nih.gov). Then a BLAST search at the International Gene Trap Consortium (IGTC) site (http://www.genetrap.org) to see if such knockouts exist.

Answer:

There are three possible knockouts. However, two of them occur at the same place.

http://www.mmrrc.org/distribution/cellLines.html

http://www.sanger.ac.uk/PostGenomics/genetrap/clones/

#7 step by step instructions

1.	Go to http://www.ncbi.nlm.nih.gov, change the Search option to Gene, enter AGPAT6 in the for box

and click Go.

2. Click on the mouse link on the search results page to go mouse Entrez Gene page. Scroll down this page to the NCBI Reference Sequences (RefSeq) section.

Click on the mRNA sequence link.

mRNA Sequence NM_018743

3. Convert the sequence in the default format to a FASTA formatted file by changing the Display option to FASTA. Copy this sequence.

4. Go to the IGTC web site (http://www.genetrap.org).

Click on DATA ACCESS in the blue navigation bar to see the options available. Select the Blast Search option.

5. Paste your mRNA sequence into the Enter sequence below box

and click Quick Search.

Using an mRNA sequence for a Blast search at this site allows the detection of standard loss-of-function allele cell lines. To find intronic ones would require the use of the sequence for the genomic region occupied by the mRNA sequence.

6. Scroll through the results and look at the actual alignments.

The IGTC web site uses strict guidelines in associating a cell line with a gene. A match needs to be at least 50% of the cell line length and have an identity of at least 90%.

Using this criteria the following three cell lines are associated with the AGPAT6 gene:

DTM030, XS0453 and XS0575
The other cell line in the top four CMHD-GT_184C11-3 does not meet this criteria.

7. Click on one of these cell line links to go off to a cell line annotation page. Here data is presented on the cell line, the gene it is associated with, and an image is given displaying the location of the cell line with respect to the gene's mRNA sequence.

To see all the data that is available on this page, click the Show All arrow in the Additional Information. To hide this information again, click on Hide All.

In the Sequence Tag Information section of the page,

information is provided on the source of the cell line and how to order it via a provided link.

In this case:

DTM030 is from BayGenomics
order from MMRRC

http://www.mmrrc.org/distribution/cellLines.html

XS0453 and XS0575 are from the Sanger International Gene Trap Resource
order from the SIGTR

http://www.sanger.ac.uk/PostGenomics/genetrap/clones/

Clicking on the Gene Description link goes off to a Gene Annotation page.

The image shows that cell lines XS0453 and XS0575 occur in approximately the same place, while DTM030 is further down stream.

last updated 4/27/2007

	and click Go.
2.	The returned Entrez page is organized with literature matches in the top box, sequence based information in the middle one and NLM's resources in the bottom box. Items of possible interest are denoted by numbers in a white box next to a topic title.

	Click on at least five diverse hits below the CFTR gene, finding out their relationship to cystic fibrosis. Ignore any gene that doesn't have a NCBI Reference Sequences (RefSeq) section.
	Here is what one of the results pages looks like.

	Click on the PUBMED links in the middle of the page and scan through the titles for mention of cystic fibrosis.

	The given example (S100A8) had 3 pages of references and the cystic fibrosis ones were on the last page.

	If no papers are listed with cystic fibrosis in the description, check out the MIM link at the top of the page or the links in the GeneOntology section.

	and click Go.
2.	Check the resulting hits to insure that the summary information on the gene mentions that various types of changes in this gene are associated with the disease.

4.	Click on the mRNA code to see the data on the actual mRNA sequence data. Scroll down the page taking in the format of the information presented. mRNA sequence link
5.	Scroll back to the top of the page and change the Display option from GenBank to FASTA. The format automatically changes. Note the difference. FASTA format is the sequence format required by many database searching programs. background information on fasta format
6.	Click back to the Entrez Gene page and repeat this process with the protein code. protein link
7.	After noting the difference, click on the NCBI logo at the top of the page.

	If results are to be displayed in a new window, click on the "Show results in a new window" box prior to clicking the BLAST button.
	Protein searches gets around the problem of multiple codons coding for the same amino acid that impacts nucleotide searches. However, depending in the information sought, this is not always possible.
9.	It may take a few seconds for the search to be completed. While waiting, click on the 7tm_1 in the image to find out about the conserved domain that was found in the sequence.

	7tm_1 indicates that the protein being search with contains the transmembrane receptor signature of the rhodopsin family of transmembrane proteins. This signature is located from residue 50 to 325 in the sequence. Close the popup window. Note the length of the query sequence, this may be given on the Query line. 413 letters Wait for the results page to appear.
	Here is a screen shot of the top part of the Blast results page.

10.	Scroll down the results page past the image with its colored horizontal bars to the Sequences producing significant alignments section.

	Scores are based on the length of the query sequence and the size of the database. Short sequences will never produce great scores. To get a E value of 0.0 requires a match of at least 330 characters. A very long sequence could easily have a match this long and still not have a match that covers a significant portion of the query sequence. Always look at the resulting alignment. The mathematics of the process can sometimes result in the strange ordering of hits.

	A hit line gives the GI number of the match sequence, the database it is from, the accession code used for the sequence, the description of the sequence from the database, its Bits score and finally the E value. Hits in the list are ordered their E value, then their Bits score which reflects the length of their actual match. Enough of the description may be given to see what species the sequence is from.

	Clicking on the link given on the left side of a hit line goes off to the actual sequence information. Clicking on the right side link moves down to the alignment data for that hit.

	Notice that there are over 30 hits with an E value of 0.0 at the top of this list and that the protein code entered is not at the top of the list. There are about 130 hits in the list which mention ADBR2, beta-2 adrenergic receptor or variations thereof before sequence description changes to something else. The first 12 hits are all from man with from 0 to 2 mismatches in the alignment. NCBI used to make an effort to remove redundant sequences, but the size of the database increased to such an extent that it was no longer possible to do this quickly enough so that it wouldn't impact the processing of new data. When an accession code begins with XP_, it means that the data is the results of an automated analysis process. This situation usually occurs when a genome sequencing project is first being analyzed. These sequences have not been checked for accuracy and can be much longer or shorter than their homologs from more mature genome studies. These sequences usually have their description start with PREDICTED:. Check out some of the hits beyond the 0.0 E values and determine where the match is actually taking place within the query sequence. PREDICTED: similar to beta-2 adrenergic receptor, [Gallus gallus] Length=397 results alignment section 12 - 396 beta-2 adrenergic receptor [Homo sapiens] Length=275 results alignment section 52- 326 beta-2 adrenergic receptor [Macaca mulatta] Length=275 results alignment section 52 - 326 beta-2 adrenergic receptor [Hylobates concolor] Length=275 results alignment section 52 - 326 beta-2 adrenergic receptor [Ateles fusciceps] Length=275 results alignment section 52 - 326 .... beta-2 adrenergic receptor [Hippopotamus amphibius] Length=275 results alignment section 52 - 326 The match is happening in the 7tm_1 region of the sequence which appears to be highly conserved.

	From the list given the structural database to use is pdb. The swissprotein database was also listed.

	Of the protein databases, swissprotein is considered to have the best annotation. One of the features they report is transmembrane segment locations when available or predicted.

3.	Change the Choose Search Set Database option from nr to swissprotein using the pull down menu,

5.	Scroll down the results to the significant alignments section and click on the sequence link containing the term ADRB2_HUMAN. It should be the first one on the list.

	link to ADRB2_human

6.	Scroll down the swissprotein data file to the FEATURES section. Then read through the listed features to find those regions called "Transmembrane region".
	The first Transmembrane region from the data file

	and then click BLAST button.
8.	Wait until results page appears.
9.	The best hit comes from Bovine Rhodopsin, but it is not very strong.

	The alignment does cover the entire area containing the transmembrane segments.

10.	Return to the main NCBI page, change All databases to Gene and enter rhodopsin homo sapiens in the for box and click Go.

	The protein code is Product NP_000530
12.	Return to the main NCBI page,

	This time on the main BLAST page, click on the "Align two sequences using BLAST (bl2seq)" link in the Specialized BLAST section.
13.	In the BLAST 2 SEQUENCES page,

2.	Choose the top hit, in this case NP_001195. Clicking on the link will take you to the page for the protein. NP_001195 data file
3.	Check out the length of the protein. The length is the second item on the "LOCUS" line.

	the protein is 1038 residues long
	(The sequence to be used in the search is 1038 residues long)
4.	To obtain the sequence of NP_001195 for the BLAST search, change the Display option of the page from GenPept to FASTA using the pull down menu.

	and then click the BLAST button.
8.	Wait until the results page appears.
9.	Check out the best hits with a description that appears to be correct,

	These results would indicate that rhesus monkey [Macaca mulatta], mouse and rat would all be good animal models in which to study this gene and it function.
10.	Return to the blastp suite submission page and change the Organism option in the Choose Search Set section from Any to Canis familiaris by clicking the Custom button and starting to enter this term into the field. Highlight the term when it appears on the list.

	For XP_536035 the alignment goes from 1 - 758 of the human sequence and the identity is 97%.
	Return to the BLAST 2 SEQUENCES page, change the code in sequence 2 box to that of the second hit

	For XP_851509 the alignment goes from 791 - 1038 and the identity is 93%.

	There appears to be about a 30 residue gap in the two dog sequences that needs to be filled before the sequence is complete.

	and click Go.
2.	Work through the list of exact gene name matches.

	From the information contained in the human Entrez Gene page, there are two known isoforms for this gene. One whose length is 997 and another whose length is 1042 (NP_733765). Use the longer one in your analysis runs.
3.	Use this RefSeq name to do a BLAST search on all vertebrate sequences, by returning to a page with the NCBI logo on it and clicking on that logo.

	and then click the BLAST button.
5.	Wait until a results page appears.
6.	Scroll down the page to the significant alignments section.

	Repeat this process with all 8 of the found swissprotein family members. You will need to scroll down to the bottom of the Enter or Paste box each time to add your new data. Be sure that there is a cursor below the // of the last entry, below this you paste in your new data.
	After all your sequences have been pasted in,

	click the Run button.
9.	The waiting page, that lets you known that the site is working on your alignment,

10a.	If the image appears, the columns are colored according to the residue type. A column that is all the same color and which contains the same character in each row is a conserved position. Care needs to be taken with the A,I,L,M,V locations, as the site considers these residues interchangeable, but they are not. Move across the alignment and see how it changes with position.
10b.	If the image doesn't appear, move down the page to the Alignment portion.

	This produces a new page where the alignment position characters are colored according to residue type.
	Again, care needs to be taken with the A,I,L,M,V locations. All the positions with an "*" in the line below the members of the alignment have the same character in that location.
11.	Depending on how careful you where in selecting sequences, there could be possible changes.
	There could be two types of proteins here based on length. Looking at the human Entrez Gene page showed that there were two isoforms of this gene. They only seem to differ at the C terminus end, where one form is longer than of other.
	The end of the alignment looks like this.

	Adding the frog data changed the aligned data. The least conserved area is at the very end, between 993 and 1044. Only the cat and dog sequences seem to short. Perhaps their sequence data is not for the same isoform as the others.
	The members of the alignment appear to be very highly conserved when the same isoform is used across species..

	Using an mRNA sequence for a Blast search at this site allows the detection of standard loss-of-function allele cell lines. To find intronic ones would require the use of the sequence for the genomic region occupied by the mRNA sequence.
6.	Scroll through the results and look at the actual alignments.

	The IGTC web site uses strict guidelines in associating a cell line with a gene. A match needs to be at least 50% of the cell line length and have an identity of at least 90%. Using this criteria the following three cell lines are associated with the AGPAT6 gene: DTM030, XS0453 and XS0575 The other cell line in the top four CMHD-GT_184C11-3 does not meet this criteria.
7.	Click on one of these cell line links to go off to a cell line annotation page. Here data is presented on the cell line, the gene it is associated with, and an image is given displaying the location of the cell line with respect to the gene's mRNA sequence.
	To see all the data that is available on this page, click the Show All arrow in the Additional Information. To hide this information again, click on Hide All.