Erratum: Corrigendum: Design of a genome-wide siRNA library using an artificial neural network

Huesken, Dieter; Lange, Joerg; Mickanin, Craig; Weiler, Jan; Asselbergs, Fred A.M.; Warner, Justin; Meloon, Brian; Engel, Sharron; Rosenberg, Avi; Cohen, Dalia; Labow, Mark; Reinhardt, Mischa; Natt, François; Hall, Jonathan

doi:10.1038/nbt1005-1315a

Cited by 4 publications

(5 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The 2431 siRNAs of length 21 nucleotides with complete base pairing to their respective target sequences from a siRNA study were used as the empirical activity data to study [18] , [19] .…”

Section: Methodsmentioning

confidence: 99%

Comparing Artificial Neural Networks, General Linear Models and Support Vector Machines in Building Predictive Models for Small Interfering RNAs

McQuisten

Peek

2009

PLoS ONE

View full text Add to dashboard Cite

BackgroundExogenous short interfering RNAs (siRNAs) induce a gene knockdown effect in cells by interacting with naturally occurring RNA processing machinery. However not all siRNAs induce this effect equally. Several heterogeneous kinds of machine learning techniques and feature sets have been applied to modeling siRNAs and their abilities to induce knockdown. There is some growing agreement to which techniques produce maximally predictive models and yet there is little consensus for methods to compare among predictive models. Also, there are few comparative studies that address what the effect of choosing learning technique, feature set or cross validation approach has on finding and discriminating among predictive models.Principal FindingsThree learning techniques were used to develop predictive models for effective siRNA sequences including Artificial Neural Networks (ANNs), General Linear Models (GLMs) and Support Vector Machines (SVMs). Five feature mapping methods were also used to generate models of siRNA activities. The 2 factors of learning technique and feature mapping were evaluated by complete 3×5 factorial ANOVA. Overall, both learning techniques and feature mapping contributed significantly to the observed variance in predictive models, but to differing degrees for precision and accuracy as well as across different kinds and levels of model cross-validation.ConclusionsThe methods presented here provide a robust statistical framework to compare among models developed under distinct learning techniques and feature sets for siRNAs. Further comparisons among current or future modeling approaches should apply these or other suitable statistically equivalent methods to critically evaluate the performance of proposed models. ANN and GLM techniques tend to be more sensitive to the inclusion of noisy features, but the SVM technique is more robust under large numbers of features for measures of model precision and accuracy. Features found to result in maximally predictive models are not consistent across learning techniques, suggesting care should be taken in the interpretation of feature relevance. In the models developed here, there are statistically differentiable combinations of learning techniques and feature mapping methods where the SVM technique under a specific combination of features significantly outperforms all the best combinations of features within the ANN and GLM techniques.

show abstract

“…The 2431 siRNAs of length 21 nucleotides with complete base pairing to their respective target sequences from a siRNA study were used as the empirical activity data to study [18] , [19] .…”

Section: Methodsmentioning

confidence: 99%

Comparing Artificial Neural Networks, General Linear Models and Support Vector Machines in Building Predictive Models for Small Interfering RNAs

McQuisten

Peek

2009

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…Many studies [5,9,10] showed that the percentage of the GC content of the siRNA play an important role in its efficiency. In Figure 3 the average GC contents of the siRNA were calculated for the three groups (high efficient, moderate efficient and low efficient), which are equals (8.8, 10, 10.9 nt/19 nt of the AS strand) respectively.…”

Section: Resultsmentioning

confidence: 99%

Selection of highly efficient small interference RNA (SiRNA) targeting mammalian genes

El-Lakkani¹,

Elgawad²,

Sayed³

2013

JBPC

View full text Add to dashboard Cite

RNAi is the method of silencing the expression of targeted genes. RNAi applications include gene function analysis and target validation. Designing highly efficient small interference RNA (siRNA) sequence with maximum target specificity for mammalian RNAi is one of important topics in recent years. In this work, a statistical analysis of the information for a large number (3734) of siRNA presented in the database available on the internet is done. This is to improve the design of efficient siRNA molecules. The (3734) siRNAs are classified according to their efficiency to three groups (high efficient, moderate efficient and low efficient). Thirteen properties (positional and thermodynamics) are identified in the high efficient group in the primary statistical study. In the final statistical study, the average weight of each identified property is calculated. A very good linear correlation was found between the average percentage efficiency and the weighted score of siRNA properties. It is found that the most important feature of highly efficient siRNA is the difference in binding energy between the 5’ end and the 3’ end of the anti-sense strand. The (RISC) activation step is a critical step in RNAi process where the efficiency of this process depends on the instability of the 5’end of the anti-sense strand.

show abstract

“…Dataset 2431 was from [23], the 21-mer sequence and activity data used was from the corrigendum [44]. Dataset 579 was from the compiled 581 19-mer sequences and activities dataset used by [25], with the exception of five sequences that did not precisely correspond to their target gene DNA sequence.…”

Section: Methodsmentioning

confidence: 99%

“…Two datasets are used in the present study. The first is a set of 2431 siRNA sequences of 21 nucleotides in length from [23], specifically from the corrigendum [44], referred to as dataset 2431 . The second is a compiled set of 579 siRNA sequences of 19 nucleotides in length from [25] referred to as dataset 579 .…”

Section: Introductionmentioning

confidence: 99%

Improving model predictions for RNA interference activities that use support vector machine regression by combining and filtering features

Peek

2007

BMC Bioinformatics

View full text Add to dashboard Cite

show abstract

Erratum: Corrigendum: Design of a genome-wide siRNA library using an artificial neural network

Cited by 4 publications

References 1 publication

Comparing Artificial Neural Networks, General Linear Models and Support Vector Machines in Building Predictive Models for Small Interfering RNAs

Comparing Artificial Neural Networks, General Linear Models and Support Vector Machines in Building Predictive Models for Small Interfering RNAs

Selection of highly efficient small interference RNA (SiRNA) targeting mammalian genes

Improving model predictions for RNA interference activities that use support vector machine regression by combining and filtering features

Contact Info

Product

Resources

About