Bioinformatics Advance Access originally published online on August 20, 2008
Bioinformatics 2008 24(20):2405-2406; doi:10.1093/bioinformatics/btn442
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
siDRM: an effective and generally applicable online siRNA design tool
Department of Neuroscience, University of Minnesota, Minneapolis, MN 55455, USA
*To whom correspondence should be addressed.
 |
ABSTRACT |
|---|
 TOP ABSTRACT 1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSIONREFERENCES |
|---|
Summary: Small interfering RNAs (siRNAs) have become an indispensable tool for the investigation of gene functions. Most existing siRNA design tools were trained on datasets assembled from confined origins, incompatible with the diverse siRNA laboratory practice to which these tools will ultimately be applied. We have performed an updated analysis using the disjunctive rule merging (DRM) approach on a large and diverse dataset compiled from siRecords, and implemented the resulting rule sets in siDRM, a new online siRNA design tool. siDRM also implements a few high-sensitivity rule sets and fast rule sets, links to siRecords, and uses several filters to check unwanted detrimental effects, including innate immune responses, cell toxic effects and off-target activities in selecting siRNAs. A performance comparison using an independent dataset indicated that siDRM outperforms 19 existing siRNA design tools in identifying effective siRNAs.
Availability: siDRM can be accessed at http://siRecords.umn.edu/siDRM/.
Contact: toli{at}biocompute.umn.edu
Supplementary information: Supplementary data are available at Bioinformatics online.
|
1 INTRODUCTION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSIONREFERENCES |
|---|
Small interfering RNAs (siRNAs) have become one of the principal tools used to investigate gene function in molecular genetics and functional genomics laboratories. Most existing siRNA design tools [see recent reviews in (Mittal, 2004; Pei and Tuschl, 2006; Swarup, 2004)] were trained on datasets assembled from very confined origins (from <10 independent studies), raising the concern that they may suffer from overfitting (Chalk et al., 2004; Saetrom and Snove, 2004). Indeed, mounting evidence indicates that the type of siRNA constructs (McManus and Sharp, 2002; McManus et al., 2002), cell lines (Elmaagacli et al., 2005; Spankuch-Schmitt et al., 2002) and gene product detection methods (Atkinson et al., 2006) strongly influences the efficacy determined for a RNAi experiment, suggesting that the training datasets used in developing siRNA design tools should reflect the vast diversity of laboratory techniques and experimental settings in the RNAi practice to curtail overfitting.
In the past three years, our effort to document all published mammalian siRNA experiments has resulted in the siRecords database (Ren et al., 2006). As the largest public database of experimentally tested siRNAs, siRecords is a faithful representation of the general, diverse siRNA experimental practice. Recently, using a training dataset compiled from siRecords (consisting of 2184 siRNA experiments originated from 1141 independent studies), we developed a bundle of highly effective siRNA design rule sets called the DRM rule sets (DRM stands for disjunctive rule merging) (Gong et al., 2006).
In this article, we present siDRM, a full-featured siRNA design tool in which several improved DRM rule sets are implemented. New features of siDRM include the incorporation of a bundle of higher sensitivity rule sets and faster rule sets, several filters to check undesirable detrimental effects and the linking to siRecords to check previously tested siRNAs for the genes of interest.
|
2 SYSTEMS AND METHODS |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSIONREFERENCES |
|---|
2.1 An updated DRM analysis
The procedure to develop the high effective DRM rule sets was described in lengthy detail in our previous report (Gong et al., 2006). An updated dataset was compiled from a recent release of siRecords database. This dataset consists of 5458 siRNA experiments targeting 2179 genes, derived from 2133 independent studies. We performed an updated survey to identify features that lead to significant improvement in the efficacy of siRNAs, and an updated feature combination analysis to identify top feature combinations (Gong et al., 2006). These feature combinations (or rules) were merged and organized using the DRM algorithm, resulting in several updated DRM rule sets. More details of these updated analyses can be found in the Supplementary Material.
2.2 High-sensitivity rule sets and fast rule sets
We attempted a strategy to boost the sensitivity of the rule sets by restricting the inclusion of features having the lower expected carrying rates in the feature combination analysis (the ‘carrying rate’ of a feature refers to the percentage of all siRNAs that carry the feature), and obtained a bundle of high-sensitivity rule sets.
We took a similar strategy to obtain a bundle of fast rule sets by restricting the inclusion of features that take long time to calculate.
2.3 Avoiding detrimental effects
The siDRM design tool is equipped with several filters that incorporate the latest developments in the mechanisms of several undesirable detrimental effects, and the ways of avoiding them.
2.3.1 Innate immune responses
siRNA duplexes can activate innate immune responses by interacting with certain toll-like receptors on the cell surface or in the endosomes (Hornung et al., 2005; Judge et al., 2005). Invoking these responses requires the presence of specific sequence motifs, 5' -UGUGU–3' or 5' -GUCCUUCAA–3' in the guide strand of siRNA duplexes. siDRM checks the occurrences of these motifs and alerts the user when they occur in candidate siRNA sequences.
2.3.2 Cell toxic effects
The presence of the motif 5' -UGGC–3' in an siRNA duplex was found to lead to strong toxic response (Fedorov et al., 2006). siDRM checks and reports the occurrences of this motif in the candidate siRNAs.
2.3.3 Off-target activities
Substantial off-target inhibition can take place when an siRNA and a non-targeted gene have exact sequence matches for all but the last two positions, which are tolerant to mismatches (Birmingham et al., 2006; Dahlgren et al., 2008). Off-target inhibition can also be induced when matches at the seed region (positions 2–7 or 2–8) are followed by several mismatched nucleotides (Birmingham et al., 2006; Jackson et al., 2006). For each candidate siRNA, siDRM checks and reports if (i) its full sequences (19-mer) has full homology to another transcript (5' UTR+CDS+3' UTR); (ii) its 17-mer subsequence (all but the last two positions) has full homology to another transcript (5' UTR+CDS+3' UTR); (iii) its seed region (positions 2–8) has full homology to the 3' UTR region of another transcript and (iv) its seed region (position 2–8) has full homology to the 3' UTR region of another transcript, and this homology region is followed by four consecutive mismatches.
2.4 Linking to siRecords
siRecords currently holds experimentally validated siRNAs targeting about 14% of the genes in the human genome and about 5% of the genes in the mouse genome (Ren et al., 2006). As the curation effort continues, there is an increasing chance that a researcher can find in siRecords siRNAs already designed and tested against the gene of his interest. The siDRM design tool is linked seamlessly to siRecords. If hits are found in siRecords for the gene of the user's interest, information about the previously tested siRNAs, including experimental settings, efficacies and links to the original publications, will be provided to the user.
|
3 IMPLEMENTATION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSIONREFERENCES |
|---|
The siDRM server was developed in PHP running under Apache 2.0 on a Fedora Core II Linux system. The Mfold package (version 3.2) (Zuker, 2003) installed on the server computer is used for local folding potential calculation.
|
4 RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSIONREFERENCES |
|---|
4.1 Utility
At the siDRM web site (http://siRecords.umn.edu/siDRM/), the user can either enter the NCBI Genbank accession number of a gene, or paste a mRNA sequence (in FASTA or plain text format) as input. The user is then provided with several rule sets options. Four ‘standard’ rule sets (RS2, RS3, RS6 and RS7), two high-sensitivity rule sets (RS_HS1 and RS_HS4) and two fast rule sets (RS_Fast2 and RS_Fast4) are implemented in siDRM. For ordinary siRNA design tasks, the user should first try one of the standard rule sets. If adequate number of siRNAs is not provided with the standard rule sets, the user can resort to the high-sensitivity rule sets, where a larger number of candidate siRNAs will be produced. The fast rule sets option should be selected if the user needs to design siRNAs for a large number of genes, and/or wants the design to be completed rapidly. When the fast rule sets option is selected, the results of the design will be displayed to the user promptly (typically within 2 min). If the standard rule sets option or high sensitivity rule sets option is selected, the user will be prompted to provide an email address and (optionally) a job name. The user is able to check the status of his or her design jobs using the email address provided, and when the design jobs are completed, retrieve the results from the server.
4.2 Performance
A performance comparison indicates that siDRM consistently outperforms 19 other siRNA design tools in identifying effective siRNAs from an independent siRNA efficacy dataset. Details of this comparison are presented in the Supplementary Material.
Conflict of Interest: none declared.
|
FOOTNOTES |
|---|
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors. 
Associate Editor: Jonathan Wren
Received on March 10, 2008; revised on August 17, 2008; accepted on August 17, 2008
|
REFERENCES |
|---|
|
TOPABSTRACT1 INTRODUCTION2 SYSTEMS AND METHODS3 IMPLEMENTATION4 RESULTS AND DISCUSSIONREFERENCES |
|---|
Atkinson PJ, et al. Altered expression of G(q/11 alpha) protein shapes mGlu1 and mGlu5 receptor-mediated single cell inositol 1,4,5-trisphosphate and Ca(2+) signaling. Mol Pharmacol (2006) 69:174–184.
Birmingham A, et al. 3' UTR seed matches, but not overall identity, are associated with RNAi off-targets. Nat. Methods (2006) 3:199–204.[CrossRef][Web of Science][Medline]
Chalk AM, et al. Improved and automated prediction of effective siRNA. Biochem. Biophys. Res. Commun. (2004) 319:264–274.[CrossRef][Web of Science][Medline]
Dahlgren C, et al. Analysis of siRNA specificity on targets with double-nucleotide mismatches. Nucleic Acids Res (2008) 36:e53.
Elmaagacli AH, et al. WT1 and BCR-ABL specific small interfering RNA have additive effects in the induction of apoptosis in leukemic cells. Haematologica (2005) 90:326–334.
Fedorov Y, et al. Off-target effects by siRNA can induce toxic phenotype. RNA (2006) 12:1188–1196.
Gong W, et al. Integrated siRNA design based on surveying of features associated with high RNAi effectiveness. BMC Bioinformatics (2006) 7:516.[CrossRef][Medline]
Hornung V, et al. Sequence-specific potent induction of IFN-alpha by short interfering RNA in plasmacytoid dendritic cells through TLR7. Nat. Med (2005) 11:263–270.[CrossRef][Web of Science][Medline]
Jackson AL, et al. Widespread siRNA "off-target" transcript silencing mediated by seed region sequence complementarity. RNA (2006) 12:1179–1187.
Judge AD, et al. Sequence-dependent stimulation of the mammalian innate immune response by synthetic siRNA. Nat. Biotechnol. (2005) 23:457–462.[CrossRef][Web of Science][Medline]
McManus MT, Sharp PA. Gene silencing in mammals by small interfering RNAs. Nat. Rev. Genet. (2002) 3:737–747.[CrossRef][Web of Science][Medline]
McManus MT, et al. Small interfering RNA-mediated gene silencing in T lymphocytes. J. Immunol. (2002) 169:5754–5760.
Mittal V. Improving the efficiency of RNA interference in mammals. Nat. Rev. Genet. (2004) 5:355–365.[Web of Science][Medline]
Pei Y, Tuschl T. On the art of identifying effective and specific siRNAs. Nat. Methods (2006) 3:670–676.[CrossRef][Web of Science][Medline]
Ren Y, et al. siRecords: an extensive database of mammalian siRNAs with efficacy ratings. Bioinformatics (2006) 22:1027–1028.
Saetrom P, Snove O. A comparison of siRNA efficacy predictors. Biochem. Biophys. Res. Commun. (2004) 321:247–253.[CrossRef][Web of Science][Medline]
Spankuch-Schmitt B, et al. Effect of RNA silencing of polo-like kinase-1 (PLK1) on apoptosis and spindle formation in human cancer cells. J. Natl Cancer Inst. (2002) 94:1863–1877.
Swarup G. How to design a highly effective siRNA. J. Biosci. (2004) 29:129–131.[CrossRef][Web of Science][Medline]
Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res (2003) 31:3406–3415.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. W. Klingelhoefer, L. Moutsianas, and C. Holmes Approximate Bayesian feature selection on a large meta-dataset offers novel insights on factors that effect siRNA potency Bioinformatics, July 1, 2009; 25(13): 1594 - 1601. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Ren, W. Gong, H. Zhou, Y. Wang, F. Xiao, and T. Li siRecords: a database of mammalian RNAi experiments and efficacies Nucleic Acids Res., January 1, 2009; 37(suppl_1): D146 - D149. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||