This work is licensed under a Creative Commons License.

1. Introduction and purpose
2. Usage
   • Specificity
   • SiRNA candidate selection and parameters for siRNA design
     • ORF target
     • Patterns
     • Pattern filtering
     • Blast filtering
     • Output
   • Output
3. How long does it takes
4. Authors and Acknowledgements
5. Before you start: If "The program doesn't work"
6. How to cite SiDE
7. Terms of use
8. Privacy and Security
9. Disclaimer
10. References

Intro and purpose

RNA interference (RNAi) is a sequence-specific post-transcriptional gene-silencing process that constitutes a powerful new tool for analyzing gene knockdown phenotypes (Fire et al., 1998). This mechanism is evolutionary conserved among eukaryotes and plays an essential role in mediating responses to exogenous RNAs (such as viruses) and in stabilizing the genome by sequestering repetitive sequences (such as transposons (Hannon, 2002, Tijsterman et al., 2002)). RNA silencing is triggered by double-stranded RNA molecules, which are cleaved by the enzyme DICER into 21- to 23-nt duplexes containing a 2-nt overhang at the 3' end of each strand. These duplexes are incorporated into a protein complex called the RNAinduced silencing complex (RISC). The RISC is responsible for the sequence-specific degradation of target RNAs that contain homologous sequences. The resulting downmodulation of the encoded proteins can subsequently lead to the induction of a specific phenotype. Here we describe a new tool for the automatic design of human and mouse siRNA, SiDE (which stands for Small interfering RNA design), that applies rules for pattern sequence selection and performs similarity searches in order to minimise crosshybridization. In addition, the presence of genomic features such as single nucleotide polymorphisms (SNPs) can be considered in the design. The tool is designed for high throughput design of siRNAs, which implies high specificity and capacity for dealing with a large number of target genes.

Usage

The main objective of the tool is the high throughput design of siRNAs. To achieve this, a programme capable of processing a large number of genes and producing lists of highly specific, putative siRNAs that, at the same time, fulfil a series of thermodynamic and empiric rules is necessary. The novelty of SIDE is the high specificity in the detection of targets for siRNA interference. Currently, SiDE can find siRNA candidates in humans and mice.

Specificity

Specificity is probably the most important factor for the successful selection of proper siRNA candidates. If the siRNA match another gene the possibility of silencing side effects cannot be discarded, and conclusions on the functional assay would be impossible. The idea of specificity comes from the goal of having a unique match in a target gene and reducing the possibility of any possible cross-hybridisation in any other gene to a minimum. Specificity is usually achieved by means of using any similarity search algorithm such as BLAST. We built up the databases for BLAST using the genes annotated in the latest NCBI build genome assembly (in the curren version, 34 for human and 32 for mouse). Sequences, unique for each gene, are obtained from Ensembl (Birney et al., 2004) and formatted (using formatdb tool) for BLAST use. Consequently, our databases (one for human and one for mouse genes) contains one single copy of each gene sequence.

SiRNA candidate selection and parameters for siRNA design

ORF target:

Patterns:

Pattern filtering:

• GC content. This filter allow the selection of siRNA with a GC content between the percentages selected. It is strongly recommended to maintain percentages between 26% and 56%, equivalent to the 30%-52% suggested by Reynolds et al., 2004.
  
  
• CT content. This filter is the same as above but for CT content.
  
  
• Avoid more than 4 (default) contigous G. Poly G stretches can form agglomerates that may interfere with the siRNA binding. With this filter you can discard patterns with polyG.
  
  
• Avoid more than 4 (default) contigous A. The same as above but for A.
  
  
• Variation of up to 20% (default) in nucleotide composition. In this way, the nucleotides do not display extreme percentage values.
  
  
• Avoid A in position 3. For example, the pattern NNAN20 would be avoided.
  
  
• After nucleotide 2 (default) set to 3 (default value in the range 2 to 4) the minimum number of G/Cs in the next 4 bases.
  
  
• After nucleotide 17 (default) set to 3 (default value in the range 2 to 4) the minimum number of A/Ts in the next 4 bases.
  
  
• Avoid SNPs. This option may be useful if the gene is suspected of being polymorphic. In this instance, silencing could be affected by mismatches due to allele variants. This option and the two to follow are active by default.
  
  
• Avoid exon boundaries. This option might be useful if alternative splicing can take place in the gene. Avoiding exon boundaries thus reduces the possibility of mispairing siRNA.
  
  
• Avoid repeat regions. Again, if repeated regions do exist it is good practice to exclude them.
  
  
• Filter oligos with less than 6 (default) points of Reynolds et al. score. This score is calculated in the same way that in Reynolds et al., 2004. There, the authors identified a set of 8 characteristics associated with siRNA functionality and developed an algorithm to predict the eficiency of a siRNA. With that algorithm a siRNA was assigned a score according to the satisfaction or not of the 8 characteristics. The score was calculated as follows:
  
  
  1. GC content between 30%-52% -> +1 point
  2. +1 point for each A/T at positions 15-19
  3. Absence of internal repeats -> +1 point
  4. An 'A' at position 19 -> +1 point
  5. An 'A' at position 3 -> +1 point
  6. A 'T' at position 10 -> +1 point
  7. A 'G' or 'C' at position 19 -> -1 point
  8. A 'G' at position 13 -> -1 point
  
  
  

  Using this algorithm most functional siRNA tested in that paper had a score of 6 or more, so a score of 6 was defined as the cutoff for selecting siRNAs.

  Here we calculate the score in the same way, using the central 19 nucleotides of the pattern (for example, if you select the pattern AAN19TT, we use the N19 to calculate the score). If you use a custom pattern, this filter will be available only if the pattern has a length of 19 bases.

  Also, to check the presence of internal repeats the mfold program (Mathews et al., 1999) was used instead of the calculation of the melting temperature of the siRNA hairpin. If the minimum folding energy is positive, the formation of internal repeats is going to be thermodynamically unfavourable, so we're goint to supose an absence of internal repeats and therefore rule number three is going to be satisfied.
  
  

• Filter oligos with a classification lower than Ib (default) based on Ui-Tei et al., 2004 classification. In this filter, the siRNA are classified according to the classification of Ui-Tei et al. 2004:
  
  
  • Class Ia: A/T at the 5' antisense (AS) end, C/G at the 5' sense (SS) end and 5-7 A/T residues in a 7 nt 5'-terminal end of AS.
  • Class Ib: A/T at the 5' AS end, C/G at the 5' SS end and 4 A/T residues in a 7 nt 5'-terminal end of AS.
  • Class II: All siRNAs that not belong either to class I or III.
  • Class III: siRNAs with opposite features with respect to class I (i.e. C/G at the 5' AS end, A/T at the 5' SS end and less than 4 A/T residues in a 7 nt 5'-terminal end of the AS).
  
  
  
  
  
• Filter oligos with less than 2 points of Amarzguioui & Prydz, 2004 score. This score is calculated as explined in Amarzguioui & Prydz, 2004:
  
  
  • A/T differential between the three terminal nucleotides in the 3' and 5' ends of the duplex -> -3/+3 points.
  • A 'C' or 'G' at position 1 -> +1 point.
  • A 'T' at position 1 -> -1 point.
  • An 'A' at position 6 -> +1 point.
  • A 'G' at position 19 -> -1 point.
  • An 'A' or 'T' at position 19 -> +1 point.
  
  
  
  
  
• Filter oligos with less than 2 points based on Hsieh et al., 2004 paper. The authors of this paper do not create specifically an algorithm to get a score for a particular siRNA, but the sequence criteria they propose can be used to construct an algorithm. Here we follow the implementation suggested in Saetrom & Snove, 2004 with an additional -1 for A/T in position 11 as suggested in Hsieh et al., 2004.
  We calculate the score as follows:
  
  
  • A 'C' at position 6 -> -1 point.
  • A 'C' or 'G' at position 11 -> +1 point.
  • An 'A' or 'T' at position 11 -> -1 point.
  • An 'A' at position 13 -> +1 point.
  • A 'G' at position 16 -> +1 point.
  • A 'G' at position 19 -> -1 point.
  • A 'T' at position 19 -> +1 point.
  
  
  
  
  
• Filter oligos with less than 10 points of Takasaki et al., 2004 score. Here the score is calculated as in Takasaki et al. 2004:
  
  
  • An 'A' at position 1 -> -3.97 points.
  • A 'T' at position 1 -> -3.75 points.
  • A 'G' at position 1 -> +7.4 points.
  • An 'A' at position 6 -> +2.33 points.
  • A 'G' at position 7 -> +2.4 points.
  • A 'T' at position 7 -> -2.59 points.
  • An 'A' at position 8 -> +3.02 points.
  • A 'G' at position 8 -> -2.35 points.
  • A 'G' at position 9 -> -2.35 points.
  • A 'T' at position 9 -> +2.3 points.
  • A 'T' at position 15 -> +2.7 points.
  • A 'G' at position 19 -> -2 points.
  
  
  

Blast filtering:

• Allow unspecific BLAST (Altschul et al., 1990) alignments with more than 4 (default in the range 1 to 10) non-identical bases (gap). This filter will include putative siRNA targets in the result if the second unspecific match (if any) has a gap of more non-identical bases than the one selected.
  
  
• BLAST alignments specific of Transcript. Ensembl facilitates the use of transcripts instead of the complete gene (in cases in which different transcripts are available), and siRNA can be selected for silencing specific transcripts if this filter is selected.

Output:

• Sorting. Results can be sorted using different criteria including all the scores, Differential Tm, number of gaps or Tm. The criteria of a number of mismatches are usually intuitive enough to establish a threshold. Nevertheless, in some situations, such as genes with biases or strong non- uniformities in base composition, mismatches can result in erroneous estimations of the likelihood of cross-hybridisations. In this instance, the application of criteria that is more related to physical likelihood of matching such as the melting temperature rather than the simple number of mismatches can be more useful. Melting temperature (Tm) of the hybrid siRNA and the target sequence is estimated by using the program MELTING (Le Novére, 2001). The differential of Tm between the specific target sequence and the next unspecific candidates can be thus used as more rational criteria to set a threshold. The scores can also be a good criteria to select the siRNAs. A comparative of the 5 scores can be found in Saetrom and Snove, 2004.
  
  
• Remove the first 2 nucleotides at 5' on Sense and last 2 nucleotides at 3' on Antisense. This option and the two next ones can be used for convenient representation of the final siRNA sequences.
  
  
• Remove the first 2 nucleotides at 5' on Sense and Antisense.
  
  
• Remove the last 2 nucleotides at 3' on Sense and Antisense.

Output

This is an example of the SiDE output in HTML format:




These are theresults for p53 (HUGO TP53) selecting the NAN19NN pattern with the default options (note that these results are based on Ensembl 22_34d, so they may not correspond to what you can have now).

The first 2 columns show the start and end positions of the designed siRNA with respect to the start of transcription of the gene.

The next columns is the pattern (or patterns) selected to make the design.

Columns four and five show the designed siRNA (sense and antisense strands respectively).

After that we have several characteristics for each siRNA like the CG% content (including the whole siRNA, not just the central N19), the melting temperature (Tm) of the sense and antisense strands of the siRNA, and the five scores implemented for that siRNA (calculated as has been seen avobe, see "Pattern filtering" section).

The salt and DNA concentration used to calculate the melting temperature are:
Sodium concentration: 20 mM
DNA concentration: 10 microM

The last 2 columns are the gap (number of mismatches) with the next best hit in the blast results, and the differential Tm of the designed siRNA and the next best hit.

How long does it takes

The time spent to predict the siRNAs is very variable. It depends on the number of genes being scanned, the number of transcripts for each gene, the length of those genes, and the options selected. The following examples might guide you in order to see how long is going to take your prediction:

Please, note that these times may have variations due to overall system loading.

Authors and Acknowledgements

This program was developped by Javier Santoyo and Juan M. Vaquerizas from the Bioinformatics Unit at CNIO under the supervision of Joaquín Dopazo. Now we're runnning the program from the Bioinformatics Department at the Centro de Investigación Príncipe Felipe (CIPF).

We want to thank the people of the CNIO Bioinformatics Unit and Cell Division and Cancer Group for beta testing, and comments on the design of the tool.

Before you start: If "The program doesn't work"

Please, before calling/emailing us if "the program doesn't work", make sure you have read this documentation. Please, also make sure that javascript is enabled on your web browser. This tool has been tested in Explorer 6.0, Netscape 7.0 and Konqueror 3.1.1.

How to cite SiDE

When using SiDE for your scientific work, please use the following reference for your publications:

Reference: Santoyo, J., Vaquerizas, J.M. & Dopazo, J. 2005 Highly specific and accurate selection of siRNAs for high-throughput functional assays. Bioinformatics, 21(8), 1376-1382.

Terms of use

• You acknowledge that the SiDE Software is experimental in nature and is supplied "AS IS", without obligation by the authors, the CIPF's Bioinformatics Department or the CIPF to provide accompanying services or support. The entire risk as to the quality and performance of the Software is with you. The CIPF and the authors expressly disclaim any and all warranties regarding the software, whether express or implied, including but not limited to warranties pertaining to merchantability or fitness for a particular purpose.

• By using results obtained with SiDE for any publication we ask you to cite the corresponding references and the main web site (http://side.bioinfo.ochoa.fib.es) in that publication.

• We ask you to give us feedback concerning bugs, errors or misconfigurations. Complaints or suggestions are welcome.

Privacy and Security

Uploaded data set are saved in temporary directories in the server and are accessible through the web until they are erased after some time. Anybody can access those directories, nevertheless the name of the directories are not trivial, thus it is not easy for a third person to access your data.

In any case, you should keep in mind that communications between the client (your computer) and the server are not encripted at all, thus it is also possible for somebody else to look at your data while you are uploading or dowloading them.

Disclaimer

This software is experimental in nature and is supplied "AS IS", without obligation by the authors or the CIPF the to provide accompanying services or support. The entire risk as to the quality and performance of the software is with you. The authors expressly disclaim any and all warranties regarding the software, whether express or implied, including but not limited to warranties pertaining to merchantability or fitness for a particular purpose.

References

• Altschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. (1990). Basic local alignment search tool. J. Mol. Biol. 215, 403-410.
• Amarzguioui, M and Prydz, H. (2004). An algorithm for selection of functional siRNA sequences. Biochem. Biophys. Res. Commun. 316, 4, 1050-1058.ç
• Asharafi, D., Chang, F.Y. Watts, J.L., Fraser, A.G. Kamath, R.S., Ahringer, J., and Ruvkun, G. (2003) Genome-wide RNAi analysis of Caenorhabditis elegans fat regulatory genes. Nature 421, 268-272
• Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschl T. (2001) Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494-498
• Elbashir, S.M. Lendeckel, W. and Tuschl, T. (2001b) RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 15, 188-200
• Hannon, G.J. (2002) RNA interference. Nature 418, 244-251
• Hsieh, A., Bo, R., Manola, J., Vazquez, F., Bare, O., Khvorova, A., Scaringe, S., Sellers, W. (2004). A library of siRNA duplexes targeting the phosphoinositide 3-kinase pathway: determinants of gene silencing for use in cell-based screens. Nucleic Acids Res. 32, 3, 893-901.
• Le Novére, N. (2001). MELTING, computing the melting temperature of nucleic acid duplex. Bioinformatics 17:1226-1227
• Mathews DH, Sabina J, Zuker M, Turner DH (1999). Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J Mol Biol. 288:911-940
• Saetrom P, Snove O Jr. (2004). A comparison of siRNA efficacy predictors. Biochem Biophys Res Commun. 321, 1, 247-53.
• Reynolds, A., Leake, D., Boese, Q., Scaringe, S., Marshall, W.S. Khvorova, A. (2004). Rational siRNA design for RNA interference. Nat. Biotechnol. 22:326-330
• Takasaki, S., Kotani, S., Konagaya, A. (2004). An effective method for selecting siRNA target sequences in mammalian cells. Cell Cycle. Epub ahead of print.
• Ui-Tei, K., Naito, Y., Takahashi, F., Haraguchi, T., Ohki-Hamazaki, H., Juni, A., Ueda, R., Saigo, K. (2004). Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference. Nucleic Acids Res. 32, 3, 936-948.

Copyright