Frequently asked questions:

What is a cDNA cluster?

What is a cDNA subcluster?

What is a cDNA clone?

What is a cDNA read?

What do the clone numbers mean?

What do the read numbers mean?

What is the meaning of the sequences?

What is the meaning of the qualities?

What steps do you take to process your data?

What is a freeze?

What does the query level do?

How do I query the annotated name?

What do the different clone annotation terms mean?

What does Machine Annotation and Curated Annotation mean?

What do the different searches mean?

How should I interpret the Gene Ontology Search page?

How should I interpret the cDNA List page?

What do the blast searches mean?

Why is the site not working correctly?

Why aren't my queries working properly?

What are the browser requirements for this site?

What BLAST databases do you search?

What parameters do you use for your BLAST searches?

Who should I contact with my questions?

How should I reference information from your site?

What is a cDNA cluster?

A cDNA cluster is a cluster of two or more cDNA subclusters that have long stretches of sequence identities, but not necessarily along the entire length of each similar cDNA subcluster. The criterion to be included in a cDNA subcluster is that greater than 95% across greater than 150 bp exists in regions of two or more subclusters. If any of the subcluster clones are less than 150 bp, then a greater than 95% identity across greater than 20 bp has to exists. A cDNA cluster with only one subcluster is a singleton cluster. In this manner, a cDNA cluster with two or more subclusters theoretically contain cDNAs, not only made from identical RNAs, but from alternatively spliced, alternatively polyadenylated, alternatively initiated RNAs from the same gene, or similar RNAs made from different but highly homologous genes.

[ top ]

What is a cDNA subcluster?

A cDNA subcluster is a cluster of two or more cDNA clones that have near sequence identities to each other. The criterion to be included in a cDNA subcluster is that greater than 99% sequence identity exists across 99% of each of the clones among each other, with a minimum of 20 bp at the ends of reads not as high homology, all due to potential sequencing errors. A subcluster with only one clone is called a singleton subcluster. In this manner, a cDNA subcluster with two or more clones theoretically contains identical cDNAs made from multiple copies of identical RNAs in the same or different tissues, from the same gene or duplicated genes in the genome.

[ top ]

What is a cDNA clone?

A cDNA clone sequence is the consensus sequence of one clone from one or more reads. In the consensus constructed clone sequence, if two reads yield different nucleotides in certain locations, the read with the highest quality in that location is selected for the consensus sequence in that location. If there are two or more clones that have identical sequence, each is listed as a separate cDNA clone consensus sequence, and not combined as is common in EST databases. By our method, one can examine small changes in alternative polyadenylation, RNA editing, and other forms of RNA processing by comparing individual clones that theoretically originate from the same gene.

[ top ]

What is a cDNA read?

A cDNA read is one sequencing reaction with a primer and specific cDNA clone template, run on a DNA sequence reader. A clone can have multiple reads that result from reads of the 3? end, 5? end, within the insert of the clone, and repeated sequence reactions of the same primer and template. Thus, reads are the raw data upon which all higher levels of sequence organization are built. We have included the reads are public access, as this will allow persons to determine quality of sequence of a particular region of interest.

[ top ]

What do the clone numbers mean?

[ top ]

What do the read numbers mean?

The clone numbers are of the format: [library][plate].[direction][sequencing number]_[well]

For example, 0033P0001Z.x0_A08 means:

Library: 0033P

Plate: 0001Z

Direction: x is 5', y is 3' (5' in this case)

Sequenceing number: first time

Well: Column A, row 8 on a 96 well plate

[ top ]

What is the meaning of the sequences?

All DNA sequences start at the read level where we call bases. We come up with a consensus sequence at the clone and subcluster levels by running phrap on the sequences below that level. When these sequences do not overlap, we add 500 N's between the sequences. We try to preserve the direction of these split sequences, so that 5' sequences come before 3' sequences. Also, we clip off primer and vector sequence from the initial reads, so this sequence does not propogate to the subcluster and cluster levels. However, we do leave any poly-A tail that we find.

[ top ]

What is the meaning of the qualities?

Calling bases at the read level produces a quality score associated with each base. This score ranges from 1 to 100. We consider 40 to be good and below 20 to be bad. The quality graphs are colored by this scale, where green is above 40, red is below 20, and in-between is black.

[ top ]

What steps do you take to process your data?

Here are the basic steps that we go through for processing and annotating:

We do base calling on reads with phred/phrap.

We clip vector and primer from each sequence, tagging the clones and reads with primer information.

We filter out poor reads (average quality < 20).

We build a consensus sequence at the clone level.

We cluster clones into subclusters and subclusters into clones, building consensus sequences/qualities at the subcluster level.

We blast all clone sequences against various public databases.

We markup the blast results with UniGene, LocusLink, and Gene Ontology annotation information.

We create annotate each level based on the marked-up blast results for each clone.

Humans curate each level based on the machine annotation and personal knowledge.

[ top ]

What is a freeze?

Whenever we add new data to the database, we freeze the old data as a stable version of the database and create a new freeze. Our freezes are numbered, starting from 1. The most recent freeze is the default freeze selected. Not all older freezes will always be available via the website, but they will always be kept on backup in case they are needed.

[ top ]

What does the query level do?

Our cDNA data is organized into four basic levels: cluster, subcluster, clone, and read.  Each level hierarchically organizes the levels below it: reads are in clones, clones are in subclusters, and subclusters are in clusters.  Curators annotate semi-independently the cluster, subcluster, and clone levels, so these are the levels that you can search at.  The Query Level selection box at the top of every screen allows you to control which of the levels you are searching at.

[ top ]

How do I query the annotated name?

You can query the annotated name via the cDNA List Search, which allows you to query  the annotated name by a full text search.  The query is case-insensitive, matches entire words, and performs an implicit AND with all the words that you enter.  An example of such a query is "c-fos".  You can also query parts of words by using the asterisk (*) character.  For example, "*fos" is a valid query.

[ top ]

What do the different clone annotation terms mean?

ORF An Open Reading Frame (ORF) is a region of a cDNA that codes for a start codon, followed by a region at least greater than 80 amino acids before reaching a stop codon.

Splice variant Two or more mRNAs that have are spliced differently from a precursor. The variants are revealed in the cDNA sequences as small to large regions (~20 bp -1Kb) that are different between the cDNAs.

Alternative adenylation When DNA polymerase synthesis mRNA, there is a poly adenylation site that signals the addition of a poly A tail to the mRNA. Some genes have multiple poly A signals in their genomic sequence such that the mRNA is either shorter or longer depending upon which poly A signal is used.

Alternative start Some genes have different promoters and/or start sites, that lead to mRNAs, and thus cDNAs that have different 5' sequences, one more upstream of the other

RNA edited After the mRNA is made from the genomic DNA sequence, individual nucleotides of some sequences are changed by RNA editing enzymes in the cell, effectively changing the RNA and the subsequent protein for those RNAs translated into a protein.

Unspliced variant This is the unprocessed RNA that was cloned as a cDNA before the cell had a chance to splice it.

Genomic This is genomic DNA that by error was cloned in the cDNA library.

Contaminant This is genomic DNA that by error was cloned in the cDNA library.

[ top ]

What does Machine Annotation and Curated Annotation mean?

Their are two ways that a cluster/subcluster/clone gets annotated: by the machine and by a curator. You will see which of these is the case at the top of the annotation. If you see Machine Annotation, then the annotation has not been looked at by an expert. If you see Curated Annotation, then the annotation has been produced by an expert with the help of the machine curation process.

[ top ]

What do the different searches mean?

The site is divided into three different ways to search the same data: the cDNA List Search, the Gene Ontology Search, and the Internal BLAST Search.  

The cDNA List Search provides a way to query the annotated names and page through those names.  You can follow links to the annotations from this view to explore all of the data associated with a given annotation and cluster.

The Gene Ontology Search provides a way to query the annotated gene ontology terms and view the number of annotations for each term.  This view is setup in a tree format, allowing you to open just the terms you are interested in.

The Internal BLAST Search provides a way to query annotations by sequence similarity.  This view allows you to enter a sequence, and then returns all of the annotations that have similar sequences and meet the current query.

[ top ]

How should I interpret the Gene Ontology Search page?

The Gene Ontology is "a dynamic controlled vocabulary that can be applied to all organisms even as knowledge of gene and protein roles in cells is accumulating and changing."  It is a hierarchical tree that becomes more specific as you get closer to the leaves.  Terms in the tree have an "is-a" or an "is-part-of" relationship with their parents.  This means that searches against a parent will always return a superset of  of the child's search results.  The Gene Ontology is divided into three major sections: molecular function, cellular component, and biological process.  All terms are part of one or more of these categories.

You can browse the Gene Ontology via our Gene Ontology Search engine.  You can open and close branches of the tree by clicking on the plus and minus signs to the left of the tree.  You can select terms for a query via the check box next to the term name.  Click on the Search button below the tree to view a cDNA List that have annotations with all of the selected terms.  Clicking on the term itself will open a new window that allows you to browse the cDNA List for just that term.  The AmiGO link brings your browser to an external site that keeps track of many of the genes that are associated with Gene Ontology terms.  Finally, the number in parentheses is the number of cDNA's annotated at or below that level.

[ top ]

How should I interpret the cDNA List page?

The cDNA List page is a listing of the cDNA's that meet your search criteria.  This criteria can include annotated name, Gene Ontology, and possibly others.  This is a listing of the cDNA's at one of the query levels that we annotate.  The first check box allows you to select specific results to be viewed later.  Click on the Save Selections button above the list to save your selections, click on the View Selections button to view the results you have saved, and click on the Clear Selections button to clear out your saved selections.  Your saved selections will be automatically purged from the system after a short period of not using the site.  Next to the check box is a link to the cluster, subcluster, or clone information for that cDNA.  From here, you will be able to access the annotation and sequence data.  Next to this link is the annotated name.

[ top ]

What do the blast searches mean?

Coming soon!

[ top ]

Why is the site not working correctly?

Why aren't my queries working properly?

What are the browser requirements for this site?

This site requires that your browser support cookies and that your browser is set to accept them.  Some of the more advanced functionality requires JavaScript and style sheets.  If your browser does not meet these requirements, the site will not perform as expected and your queries will not be processed properly.  We recommend Netscape 6+ or Internet Explorer 6+.  Minimal functionality should be available for Netscape 4+ and Internet Explorer 4+.

[ top ]

What BLAST databases do you search?

Freeze Databases

1 GenBank nt

2 GenBank nt, TIGR chicken index

3 GenBank nt, TIGR chicken index

[ top ]

What parameters do you use for your BLAST searches?

Freeze Parameters

1 none

2 none

3 -e 1e-2

4 -e 1e-2

[ top ]

Who should I contact with my questions?

General information about the annotation process Erich Jarvis jarvis@neuro.duke.edu PI

Information about the web interface Patrick McConnell mccon012@mc.duke.edu primary software developer

Clone requests None yet available  

[ top ]

How should I reference information from your site?

E. D. Jarvis, V. A. Smith, K. Wada, M. V. Rivas, M. McElroy, T. V. Smulders, P. Carninci, Y. Hayashizaki, F. Dietrich, X. Wu, P. McConnell, J. Yu, P. P. Wang, A. J. Hartemink, S. Lin. A framework for integrating the songbird brain. J Comp Physiol A (2002) 188: 961-980.

[ top ]

Collaboration

We thrive on rich collaborations to push our thinking. A continuous state of reinvention, driven by our partners in the process, is essential to our work.

Sustainability

Sustainability is at the heart of our design approach. We audit projects against global standards as well as our own, more comprehensive, responsibility framework.