Design Principles for RNAi · Design Principles for RNAi Shifra Ben-Dor Bioinformatics and Biological Computing Unit Weizmann Institute of Science COBI - INN

Design Principles for RNAi

Shifra Ben-DorBioinformatics and Biological Computing Unit

Weizmann Institute of Science COBI - INN

Optimally, we would like to take our genesequence, plug it into a program, and comeout with one good candidate that will knockdown our gene by more than 90%.

In reality, this doesn’t work.

Not all siRNAs work with equal efficiency.

If we choose sequence at random, there isgood chance we can get some knockdown

So why is it so hard to find agood siRNA?

• Cellular Factors

• Molecular Factors

• Sequence Based Factors

Pancoska et al. NAR 2004 32(4):1469-1479

Cellular Factors• Species Differences• Cell Lines/Types• Transfection Efficiency of siRNA

– Subcellular localization

• Expression Level of Target– Endogenous - low may be problematic– Exogenous - high may be problematic

Molecular Factors

• Degradation• Accidental activation of IFN response• Non-specific binding• Cross-hybridization

Sequence Based Factors

• Base Preferences• Structural Constraints• Thermodynamic Constraints

SFold http://sfold.wadsworth.org/sirna.plWhitehead siRNA Selection Web Server http://jura.wi.mit.edu/siRNAext/siDirect http://design.rnai.jp/sidirect/index.php (Ui-Tei)DEQOR http://cluster-1.mpi-cbg.de/Deqor/deqor.htmlAmbion http://www.ambion.com/techlib/misc/siRNA_finder.htmlDharmacon http://design.dharmacon.com/Emboss http://inn.weizmann.ac.il/EMBOSS/ Under Nucleic composition: sirnaJack Lin's siRNA finder: http://www.sinc.sunysb.edu/Stu/shilin/rnai.htmlOptiRNAi http://bioit.dbi.udel.edu/rnai/Filtering ineffective http://i.cs.hku.hk/~sirna/software/sirna.phpQiagen (Xeragon) http://www1.qiagen.com/Products/GeneSilencing/CustomSiRna/SiRnaDesigner.aspxInteragon http://www.interagon.com/demo/Hannon Lab http://katahdin.cshl.org:9331/RNAi_web/scripts/main2.plWistar: http://hydra1.wistar.upenn.edu/Projects/siRNA/siRNAindex.htmOligoEngine http://www.oligoengine.com/TROD http://websoft2.unige.ch/sciences/biologie/bicel/RNAi.html (t7)IDTDNA http://biotools.idtdna.com/rnai/Promega http://www.promega.com/siRNADesigner/program/GenScript https://www.genscript.com/ssl-bin/app/rnaisiSearch http://sonnhammer.cgb.ki.se/siSearch/siSearch_1.6.htmlsiRNA wizard http://www.sirnawizard.com/design_advanced.phpClontech http://bioinfo2.clontech.com/rnaidesigner/

“Rules”

• Tuschl• Reynolds• Stockholm (Sonnhammer)• Ui-Tei• Amarzguioui• And many more….

Basics

• Length of 21 base pairs total• Length of 19 base pairs• 3’ Overhang of 2 base pairs• Limit G/C content• Limit consecutive stretches of the

same base

Tuschl Rules

• Look for AA(N19)TT• ~50% G/C• If not found NA(N21)• TT was chosen for the overhang to

simplify chemical synthesis• If you use PolIII start with a purine,

no polyA tracts

Factors we can use

• Base Preferences• Thermodynamic Constraints• Structural Constraints• Minimize cross-hybridization

Taken from:Saetrom and Snove BBRC 321:247-253 (2004)

General Summary - Bases• Avoid more than 3 consecutive same

bases• Avoid G/C stretches

– 7 or more in a row– ~30-50% G/C overall

• A/U at the 5’ anti-sense is good• G/C at the 3’ anti-sense is good• No Mismatches

Mismatches

• In the middle are not tolerated (won’t getknockdown)

• At the ends can be tolerated• May switch to the miRNA path

(translation stop, not transcriptdegradation).

• Watch out for SNPs

“Thermodynamic” Constraints

• G/C content• Hairpin formation (of the siRNA)• Asymmetric Ends

– Low internal stability on 5’ of anti-sense– Differential between 5’ of sense and anti-

sense• Positions 9-14 of anti-sense - low stability

(position 10 of target=cleavage site)

Thermodynamic Constraints

Measured with:

• Melting Temperature• INN (individual nearest neighbor)

– Base pairing– Stacking

“Structural” Constraints

• Target secondary structure– Loops vs Double stranded regions

• siRNA secondary structure• UTRs may have binding factors• ATG might have binding factors

Minimizing cross-hybridization

• Check against genome database• Check against transcript database• Mismatches in the middle are preferred• Minimize consecutive matching bases• Don’t forget to check for splice variants

Taken from:Saetrom and Snove BBRC 321:247-253 (2004)

Problems comparing methods

• Different Design Strategies• Different measures of knockdown

efficiency– RNA– Protein

• Different measures of algorithmefficiency

Problems comparing methods

• Take from different parts of mRNA– CDS– UTR

• Some score and rank, others don’t• Different databases

– Genomic– mRNA/EST/Unigene

SFold http://sfold.wadsworth.org/sirna.plWhitehead siRNA Selection Web Server http://jura.wi.mit.edu/siRNAext/siDirect http://design.rnai.jp/sidirect/index.php (Ui-Tei)DEQOR http://cluster-1.mpi-cbg.de/Deqor/deqor.htmlAmbion http://www.ambion.com/techlib/misc/siRNA_finder.htmlDharmacon http://design.dharmacon.com/Emboss http://inn.weizmann.ac.il/EMBOSS/ Under Nucleic composition: sirnaJack Lin's siRNA finder: http://www.sinc.sunysb.edu/Stu/shilin/rnai.htmlOptiRNAi http://bioit.dbi.udel.edu/rnai/Filtering ineffective http://i.cs.hku.hk/~sirna/software/sirna.phpQiagen (Xeragon) http://www1.qiagen.com/Products/GeneSilencing/CustomSiRna/SiRnaDesigner.aspxInteragon http://www.interagon.com/demo/Hannon Lab http://katahdin.cshl.org:9331/RNAi_web/scripts/main2.plWistar: http://hydra1.wistar.upenn.edu/Projects/siRNA/siRNAindex.htmOligoEngine http://www.oligoengine.com/TROD http://websoft2.unige.ch/sciences/biologie/bicel/RNAi.html (t7)IDTDNA http://biotools.idtdna.com/rnai/Promega http://www.promega.com/siRNADesigner/program/GenScript https://www.genscript.com/ssl-bin/app/rnaisiSearch http://sonnhammer.cgb.ki.se/siSearch/siSearch_1.6.htmlsiRNA wizard http://www.sirnawizard.com/design_advanced.phpClontech http://bioinfo2.clontech.com/rnaidesigner/