8-Oxoguanosine switches modulate the activity of alkylated siRNAs by controlling steric effects in the major versus minor grooves.

Small interfering double-stranded RNAs have been synthesized bearing one or more base modifications at nucleotide positions 4, 11, and/or 16 in the guide strand. The chemically modified base is an N(2)-alkyl-8-oxo-7,8-dihydroguanine (alkyl = propyl, benzyl) that can alternatively pair in a Watson-Crick sense opposite cytosine (C) or as a Hoogsteen pair opposite adenine (A). Cellular delivery with C opposite led to effective targeting of A-containing but not C-containing mRNA sequences in a dual luciferase assay with RNA interference levels that were generally as good as or better than unmodified sequences. The higher activity is ascribed to an inhibitory effect of the alkyl group projecting into the minor groove of double-stranded RNA preventing off-target binding to proteins such as PKR (RNA-activated protein kinase).

Synthesis of N2-alkyl-8-oxo-7,8-dihydro-2'-deoxyguanosine derivatives and effects of these modifications on RNA duplex stability.

N(2)-alkyl analogues of 8-oxo-7,8-dihydro-2'-deoxyguanosine (OG) were synthesized (alkyl = propyl, benzyl) via reductive amination of the protected OG nucleoside and incorporated into various positions of an RNA strand. Thermal stability studies of duplexes containing A or C opposite a single modified base revealed only moderate destabilization. Both OG as well as its N(2)-alkyl analogues can pair opposite A or C with nearly equal stability, potentially offering a new means of modulating RNA-protein interactions in the minor vs major grooves.

Chemical modification of siRNA bases to probe and enhance RNA interference.

Considerable attention has focused on the use of alternatives to the native ribose and phosphate backbone of small interfering RNAs for therapeutic applications of the RNA interference pathway. In this synopsis, we highlight the less common chemical modifications, namely, those of the RNA nucleobases. Base modifications have the potential to lend insight into the mechanism of gene silencing and to lead to novel methods to overcome off-target effects that arise due to deleterious protein binding or mis-targeting of mRNA.

Copper/H2O2-mediated oxidation of 2'-deoxyguanosine in the presence of 2-naphthol leads to the formation of two distinct isomeric adducts.

Exposure of cells to phenolic compounds through exogenous and endogenous sources can lead to deleterious effects via nucleobase modifications of DNA occurring under oxidative conditions. 2'-Deoxyguanosine (dG) is the most electron rich of the four canonical bases and includes many nucleophilic sites; it is also susceptible to oxidation with numerous reactive oxygen species. In these studies, dG was allowed to react with 2-naphthol in the presence of copper or iron salts yielding two principal isomeric products. Spectroscopic analysis and reactions with alkylated nucleosides support the assignment of compound 1a/1b as a pair of atropisomer N(2) adducts and compound 2a/2b as a diastereomeric mixture of tricyclic [4.3.3.0] adducts. Both products are the result of an overall four-electron oxidation process and consequently have the same masses, though drastically different structures, providing mechanistic insight into their formation. Thus, dG alkylation by 2-naphthol under oxidative conditions yields products whose structural properties are altered, leading to potentially mutagenic effects in genomic DNA.

Leveraging voxel-wise segmentation uncertainty to improve reliability in assessment of paediatric dysplasia of the hip.

PURPOSE: Estimating uncertainty in predictions made by neural networks is critically important for increasing the trust medical experts have in automatic data analysis results. In segmentation tasks, quantifying levels of confidence can provide meaningful additional information to aid clinical decision making. In recent work, we proposed an interpretable uncertainty measure to aid clinicians in assessing the reliability of developmental dysplasia of the hip metrics measured from 3D ultrasound screening scans, as well as that of the US scan itself. In this work, we propose a technique to quantify confidence in the associated segmentation process that incorporates voxel-wise uncertainty into the binary loss function used in the training regime, which encourages the network to concentrate its training effort on its least certain predictions. METHODS: We propose using a Bayesian-based technique to quantify 3D segmentation uncertainty by modifying the loss function within an encoder-decoder type voxel labeling deep network. By appending a voxel-wise uncertainty measure, our modified loss helps the network improve prediction uncertainty for voxels that are harder to train. We validate our approach by training a Bayesian 3D U-Net with the proposed modified loss function on a dataset comprising 92 clinical 3D US neonate scans and test on a separate hold-out dataset of 24 patients. RESULTS: Quantitatively, we show that the Dice score of ilium and acetabulum segmentation improves by 5% when trained with our proposed voxel-wise uncertainty loss compared to training with standard cross-entropy loss. Qualitatively, we further demonstrate how our modified loss function results in meaningful reduction of voxel-wise segmentation uncertainty estimates, with the network making more confident accurate predictions. CONCLUSION: We proposed a Bayesian technique to encode voxel-wise segmentation uncertainty information into deep neural network optimization, and demonstrated how it can be leveraged into meaningful confidence measures to improve the model's predictive performance.