For sequences designed for use in human cells, ≤ 50 is ideal. Regardless of POTS, use proper controls to ensure effects are specific to your intended target knockdown.
Sequence can be either DNA or RNA and submitted as plain text or FASTA format.
In our hands, in the context of miRNA/shRNA expression vectors, designing 4 sequences typically results in at least one effective sequence. That being said, our program does not score or rank potency, so if you want to see close to 100% knockdown of your target, either look to the literature or other websites for additional potency criteria or design more sequences.
The RNAi design tool is limited 100kb of input sequence. The tool to evaluate siRNA sequences is limited to 10.
The lowercase letters represent G:U wobbles introduced into the siRNA sequence to improve loading of the guide strand into the RNAi effector complex, RISC, while reducing/eliminating loading of the passenger strand. If the passenger strand is loaded, it too can impart off-target effects. Northern blots to detect the passenger strand, or reporter constructs with artificial binding sites can be used to assess passenger strand loading if it is suspected.
This means that these two bases can vary and don't depend on the target sequence. In the context of siRNA design, these bases are classically chimeric DNA bases, dTdT. However, if these duplexes will be put in the context of a miRNA or shRNA shuttle, these nucleotides should match those that are typically at this position in the shuttle of choice.
Many microRNAs (miRNAs) are highly conserved with conserved target sites in mRNA 3'UTRs. Target site conservation tends to the most predictive factor of site functionality. These columns show whether the siRNA's seed sequence matches the seed sequence of a known miRNA. H(uman), M(ouse) and D(og) in the conservation column give a sense of how conserved the miRNA is. H, M or D alone means that the miRNA is restricted to a particular species or that lineage. H, M on the other hand means that orthologus miRNAs are found in both human and mouse. H, M and D mean that the microRNA is at least conserved between those species, and are often much more conserved than that. The more conserved, the more likely they have conserved, functional sites. We suggest avoiding siRNAs with miRNA seed sequences, because it is likely that the binding sites have a higher probability of being off-targeted than our calculations predict.
POTS is calculated based on known miRNA targeting rules. Three primary site types are commonly described that are more predictive of efficacy than minimal 6mer match: (in decreasing order of potency) 8mer>7mer-M8>7mer-1A>>6mer. Using available microarray data for siRNA expression studies, we predicted site types for each siRNA and calculated the probability that a transcript with a single site of any type would be down-regulated at least 0.3 fold on a log2 scale. The average probabilities for each site type were applied to all 7mer/target pairs to estimate the number of transcripts that would actually be down regulated by the siRNA. The "More Info" link on the output page has an option to show the summaries for each site type that contributed to the POTS value.
POTS cannot determine whether a particular siRNA will be toxic or cause "noticeable" off-target effects. By choosing sequences with lower POTS, you simply decrease the probability of down regulating a large number of transcripts and/or decrease the likelihood of hitting an "important" transcript(s). If you suspect seed-mediated off-targeting is an issue, either design more sequences, or create a seed-match control. A seed-match control has the same seed as the siRNA, but has mismatches (usually in the middle around bases 10-11) to the intended target. This control will not repress the intended target, but should maintain a similar off-targeting profile. Alternatively, do a rescue experiment to add back a mutated version of your target transcript that cannot be regulated by the siRNA. If either results in rescue of your toxic/undesired phenotype, your problem is not seed-mediated off targeting. If your siRNA with a high POTS score does not appear to affect your cells - great! Still, proper controls, such as those described above, will ensure that any results you see are actually due to knock down of your intended target.
50 should not be thought of as a strict cutoff. If you need to go higher, test just as you would sequences with lower POTS. If you include proper controls, find no toxicity and your results seem specific to knockdown of your target gene, you can feel confident with your results. Alternatively, decrease G/C's required at 5' end of passenger or widen slightly G/C% to loosen the potency filtering options further. Additional testing will need to be done to ensure proper strand loading and efficacy.I got back dozens of candidate siRNAs with POTS < 50. How can I shrink my list?
The actual POTS will differ slightly between species simply because of differences in the number and length of 3'UTR sequences. Based on POTS comparison between human and mouse, the rank-order of the seeds do not significantly change. In other words, a low POTS in mouse tends to be low in human. That being said, previous research has shown that siRNA off-targeting is species-specific. This means that while the overall POTS is not likely to change drastically, the specific transcripts that actually change due to off-targeting will likely be different.
We have calculated tissue-specific POTS based on expression data from 177 different human tissues and cell lines, using 3'UTR sequences only for the transcripts present in each tissue. Because POTS is based on the number of target sites, the POTS value differs in each tissue depending on the number of expressed transcripts; however, there was no significant change in POTS rank order. In other words, the order of siRNAs ranked by POTS using all 3'UTR sequences varied little with the order using only the 3'UTRs present in a specific tissue. That being said, we acknowledge that in an experimental setting the specific off-targeted genes, and therefore the off-target effects, will likely be different. The provided lists of all off-targets can be searched for particular genes of concern.
Yes. Clicking the "More Info" tab on the output page will bring up a menu with several options, one of which is a list of off-targets. The genes most likely to be down-regulated are shown first. The list can be sorted and searched on the site, or can be exported to a CSV file which can be opened in programs like Microsoft Excel.
The 18mer represents the "core" sequence that is often used for this purpose. Studies have suggested that positions 1, 20 and 21 of the antisense strand have little impact on silencing efficacy when mismatches to the intended target are introduced. The 18-mer is best to use when searching for off-targets, so mismatches to the terminal nucleotides are not counted against it. The 21-mer may be better to use to find "on-targets" - other transcript isoforms, close paralogs, or the same gene in another species.