Go to RNAiDB Home Adult Wild-type C. elegans Movie Wild-type C. elegans Movie Wild-type C. elegans Movie Wild-type C. elegans Movie
 


I. Searching RNAiDB

The RNAiDB Home Page provides a Basic Search or Quick Phenotype Search.
The Search Index Page provides more search options based on phenotypes or other features (e.g. chromosomal location or keywords/text), or using the Ace commandline. Each of these options is described below.

A. Phenotype Searches

A.1. Quick Phenotype Search (RNAiDB Home Page)

Search for RNAi Experiments based on one or more phenotypes:

1. Select a phenotype from the drop-down Query menu (all phenotypes annotated in RNAiDB are listed here).

2. Optionally, further limit the search based on whether experiments producing the selected phenotype also DO or DO NOT produce one or more of: Lethality, Sterility, or Postembryonic phenotypes. Lethality can be further limited by life-stage (embryonic or postembryonic), and Sterility can be limited further to Ste (sterility of the treated animal, or "P0 sterile") or Stp (sterile progeny).

Some examples:

  • To find a list of experiments that produced abnormal spindles in the early embryo ("Abs") and caused embryonic lethality, select "Abs" from the Query menu and "Emb" from the Lethality menu.
  • To find experiments that were "Abs" and caused sterility, select instead "Any Sterility" from the Sterility menu.
  • If you are interested only in those experiments with the "Abs" phenotype that also produced other postembryonic phenotypes, select "Yes" from the PostEmb menu.
  • All three of the above options can be combined by selecting them all of them together.


A.2. Simple Phenotype Search (available from the Search Page)

Search for RNAi experiments that produce a combination of any number of phenotypes. Phenotypes are presented in a list segregated by life-stage (Embryonic, Larval, or Adult). To select a phenotype, check the box next to the phenotype of interest.

Any number of phenotypes can be selected at the same time. If multiple phenotypes are selected, they can be combined using a logical AND or OR operator found at the bottom of the page (default is "AND").


A.3. Advanced Phenotype Search (available from the Search Page)

Search for RNAi experiments displaying a particular combination of phenotypes using a query builder form. The search form presents several rows of phenotypes that can be selected from a menu and combined with logical operators. Once the query is constructed, you will have the option to edit the query using parentheses to create a selection hierarchy, or otherwise modify it (e.g. by changing the logical operators or phenotypes to be included in the search).

To construct a query:
1. Choose a phenotype from the FIRST menu by highlighting it with the cursor.
2. Choose a selector from the drop-down menu to the LEFT (default is '=', but many other options are available).
3. Choose a logical operator from the drop-down menu to the RIGHT (default is 'END').

Optional:
4. Repeat steps 1-3 for the second and third menus.
5. If more phenotypes are desired, select the number of additional menu lines to add and press the "Add" button.

PLEASE NOTE: the query builder will stop at the first occurrence of the 'END' operator. Therefore, to select multiple phenotypes, make sure that only the last operator at the end of the form is 'END'.

6. When you are through, press the "Submit query" button. A new page displaying the query you have constructed will appear. At this point, the query can be modified to create a hierarchy of logical operators.

For example, if the original query is:
   Find RNAi Phenotype="Emb" AND Phenotype="Bmd" OR Phenotype="Ste" AND Phenotype="Unc"
It is possible to create several logically different queries by placing parenthesis in different places:

  • Find RNAi (Phenotype="Emb" AND Phenotype="Bmd") OR (Phenotype="Ste" AND Phenotype="Unc") ...
    ... will find all experiments with both Emb and Bmd phenotypes, or else both Ste and Unc phenotypes.
  • Find RNAi Phenotype="Emb" AND (Phenotype="Bmd" OR Phenotype="Ste") AND Phenotype="Unc" ...
    ... will find all experiments with Emb and Unc phenotypes, and either Bmd or Ste phenotypes.
  • Find RNAi Phenotype="Emb" AND (Phenotype="Bmd" OR (Phenotype="Ste" AND Phenotype="Unc")) ...
    ... will find all experiments that are Emb, and either Bmd or both Ste and Unc.

Thus, placement of parentheses clarifies the desired outcome of the query, so pay attention! Once you are satisfied with the query, press "Submit query" again to see the final result.


A.4. PhenoBlast (available from the Search Page)

PhenoBlast ranks RNAi experiments in the database on the basis of their overall phenotypic similarity to a single query, which may be either a Gene/Protein or an RNAi Experiment. PhenoBlast uses phenotypic signatures, i.e. a discrete set of phenotypic characters systematically scored for each experiment, to make the comparisons. (See "Phenotypic Signatures" below for more details.)

The "PhenoBlast Score" provides a measure of the likelihood of observing a particular combination of matches in the population. The PhenoBlast Score is based on the idea that matches between phenotypes which appear rarely in the population are more significant than matches between common phenotypes because they are less likely to occur by chance, and therefore should score higher.

Links to run PhenoBlast automatically are present from all RNAi Experiment and Gene Card pages for which a phenotypic signature has been scored. To perform a manual PhenoBlast search:

  • Select either Gene, RNAi Experiment, or User-defined signature.
  • In the query text box:
    A) For Gene or RNAi: Enter the name of the corresponding object.
    B) For User-defined query: Paste a signature score (as a string of numerical values) into the text box.
    (See "User-defined Signatures" below for more details.)
  • Select a phenotypic signature (currently Piano_EE, GrossPheno, or Cenix_EE).
    (See "Phenotypic Signatures" below for more details.)
  • A color palette may also be selected for signatures with multivalued characters (at present, only the Cenix_EE signature).
  • Select the number of experiments to view (default is 10).
  • Press the PhenoBlast button, and wait a few seconds for the results to appear.

The PhenoBlast Result Table provides a ranked list of the most similar signatures in the database to the query (from top to bottom, with the query itself appearing in the top row). Each row in the table corresponds to a phenotypic signature in the database, and provides the name of the RNAi Experiment, the corresponding Gene(s), a graphical representation of the signature (a key is provided at the top of the table), and a set of match scores.

PhenoBlast Ranking and Match Scores

Summary
PhenoBlast ranks results based on the sum across all characters of observed matches and the cumulative probability of the observed matches at each character position. For each pair of signatures, PhenoBlast calculates several statistics, including the number of matching defective (non-WT) characters, the number of all (non-WT and WT) matching characters, the negative sum of log probabilities of observed matches (across non-WT characters or across all characters), and a relative log probability score representing the ratio of the observed score to the score for a perfect (self) match. For multi-valued signatures, near/partial matches as well as perfect matches are considered. For display purposes, results are ordered using a lexicographic sort of the primary statistics.

Detail
For each pair of signatures A (the query signature) and B (the test signature), and phenotypic characters i (ranging from 1 to n, the length of the phenotypic signature), the PhenoBlast Score is calculated as the negative sum across all phenotypic characters i (i=1 to n) of the log probability of a match between A and B at each character position i:
      PhenoBlast Score = -Σi matchiAB*log(pi(matchAB))
MatchiAB=1 when both A and B are non-zero (i.e. not WT) at position i. The match probability pi(matchAB) = w*freq(Ai)*freq(Bi), where freq(Ai) and freq(Bi) represent the frequency of appearance of the observed character states Ai and Bi for phenotype i in the population of signatures. Thus pi(matchAB) contributes to the sum for each character position i at which the query and test signatures have a match or near match (otherwise matchiAB=0). The weight factor w=1 for matches in binary signatures. For multivalued signatures, a predetermined weight matrix is used to weigh partial matches vs. perfect matches (which receive the maximal weight). Thus matches between character states that are rare in the population influence the score more strongly than common character states, and perfect matches impact the score more strongly than partial matches. A relative log score is also calculated for the current score relative to a perfect (self) match score.

All of the above statistics are shown in the results page, and are ordered for display using a lexicographic sort on the various statistics. For binary signatures, the default sort uses first the number of matching defects, then the number of matching wild-type characters, and then the probability of observing the particular combination of matches that exists. For multivalued signatures, by default results are sorted using the total number of perfect plus partial matches followed by the match probability.

Some signatures contain character values that can't be compared (e.g. "not determined" if the character was ambiguous, or "not applicable" if the character could not be observed, as may occur if the animal dies before the stage when that character becomes manifest). These characters are excluded from the match calculations when present in the query signature.

Phenotypic Signatures

Phenotypic signatures represent a set of systematically scored discrete phenotypes, or "phenotypic characters". When a phenotypie signature is present for a particular RNAi experiment, that experiment has been explicitly scored for the presence or absence of each phenotypic character in the corresponding signature. RNAiDB currently maintains data for three phenotypic signatures:

  • Piano Early Embryonic (Piano_EE): 47 high-resolution cell biological phenotypes in the early embryo up to the 4-cell stage, based on DIC time-lapse movies, as described in Piano et al., 2002.
    This is a binary signature, i.e. each character was scored as "present" or "absent". Two additional character states were also scored: "could not be determined" (data not sufficient to call the phenotype either way), and "not applicable" (e.g. the embryo died before that character could be scored).

  • Cenix Early Embryonic (Cenix_EE): 45 high-resolution cell biological phenotypes in the early embryo through the 2-cell stage, as described in Sönnichsen et al., 2005.
    This is a multivalued signature, meaning that each character scored as "present" was also assigned a reproducibility or penetrance based on the number of observed instances relative to the total number of embryos observed.

  • Gross Phenotype (GrossPheno): This binary signature of 37 gross morphological phenotypes was auto-generated by RNAiDB for all experiments from large-scale studies for which this type of dat was collected. Examples of such phenotypes are Adl, Age, Bli, Bmd, Clr, Dpy, Egl, Emb, etc.
    Please note: While the presence of a pheotypic signature implies that all phenotypes contained in the signature were explicitly scored, theis may not necessarily be th ecase because most studies did not report when a particular phenotype was examined but not observed. Thus the absence of any given phenotypic character for this signature cannot always be interpreted as a definitive result.

User-defined Signatures

If you have scored an experiment using one of the phenotypic signatures above, you can compare your results directly to those in the RNAiDB database. A web-based scoring form for generating a phenotypic signature online is not currently available; in the meantime it is possible to enter a numerical representation of a phenotypic signature in the text box, consisting of a sequence of valid character states of the correct length (follow hyperlinks to see a list of phenotypic characters for each signature):

  • Piano Early Embryonic (Piano_EE):
    Number of characters: 47
    Character states: 0=WT ; 1=defective ; 2=not determined ; 3=not applicable
       Note: character states 2 and 3 are not used for PhenoBlast comparisons.
    Example: 00000000000000000000000001000000000111331001010

  • Cenix Early Embryonic (Cenix_EE):
    Number of characters: 45
    Character states: 0=WT ; 1=5-10% ; 2=10-20% ; 3=20-40% ; 4=40-60% ; 5=60-80% ; 6=80-100%
    Example: 001010000100000020011100100010010001000010000

  • Gross Phenotype (GrossPheno):
    Number of characters: 37
    Character states: 0=WT ; 1=defective
    Example: 0000000100110000000000000000000000010



B. Other Searches

B.1. Basic Search (RNAiDB Home Page)

Search for RNAi Experiments directly associated with other pieces of data: genes or proteins, phenotypes, GO terms, or other named objects such as RNAi Experiments, RNAi reagents (PCR product or cDNA clone), or Movies.

To use this search, go to the "Basic Search" box, select an object class, and type a name or keyword into the text box. Some examples:

  • To find all experiments associated with par-3, select Gene/Protein from the drop-down menu, and enter "par-3" into the text-field.
  • To find a set of RNAi Experiments associated with phenotypes involving cell division, select Phenotype from the drop-down menu, and enter "*division*" into the text field.
    => Note use of asterisks (the 'wild-card') to match any characters. A full list of phenotypes is available in the Quick Phenotype Search drop-down menu.
  • To see all RNAi experiments that have a movie assiocated with them, select Movie from the drop-down menu, and enter "*" into the text field.


B.2. Simple Object Search (available from the Search Page)

Search for any object stored in the database by name or partial name match.

Select an object type from the radio buttom menu, or just use the default "Anything" option to search all objects in the list.
Enter a name into the text box, optionally using wild-cards ('?' for any single character; '*' for zero or more characters).



B.3. Location Search (available from the Search Page)

Search for RNAi experiments by genomic coordinates or with a specified genomic interval. This search is useful for identifying RNAi results mapping to a particular chromosome region, and can thus help identify partially mapped candidate genes for mutational screens.
  • Select a chromosome from the drop-down menu and specify numerical coordinates in one or both text boxes.
    NOTE: If either box is left empty, a 10kb interval around the coordinate provided will be searched by default.
     
  • Select a view for the results:
    => The RNAiDB Search Results Page will display a list of experiments, genes inhibited, and phenotype summaries for all experiments found within the specified interval.
    => GBrowse will display a coordinate-based map of all gene models and corresponding RNAi Reagent mapping within the specified interval.


B.4. Text Search (available from the Search Page)

Search for any keyword or text occurring in the database. Wild-cards are optional ('?' for any single character; '*' for zero or more characters).

Type the query in the text box and select either a fast or in-depth search. The fast search will look for any objects that match the query string, as well as tags and short text fields within them. The in-depth search will additionally search all long text fields. For more information about AceDB data models and queries, please refer to the AceDB Documentation at www.acedb.org.



B.5. Ace Query (available from the Search Page)

Search RNAiDB using a command line interface to AceDB. This allows users to execute customized searches that cannot be performed using the other search options.

For more information about the Ace query language and examples, please consult the AceDB Documentation at www.acedb.org.



II. RNAi and Gene Card Pages



Help documentation for these pages is not yet finished but will be completed as soon as possible.
We thank you for your patience.

A. RNAi Experiments

B. RNAi Inhibition

1. ePCR

2. Sliding n-mer Window

3. BLAST

C. Phenotypes

1. Phenotypes by Life Stage

2. Images, Stacks, & Movies

3. PhenoBlast

D. Gene Details

1. Homology Group & KOGs

2. GO Terms

E. Gene Graphs

1. Gene Model

2. RNAi Reagent: Genomic Location

3. RNAi: Sliding Window Mapping (Reagent to Target Transcript)

4. RNAi: Sliding Window Mapping (Reagent to NON-target Transcript)

5. RNAi: BLAST Mapping

6. Transcript: Similarity with Other Genes (Sliding Window)