Transcriptome-Wide Off-Target Effects of Steric-Blocking Oligonucleotides.

Steric-blocking oligonucleotides (SBOs) are short, single-stranded nucleic acids designed to modulate gene expression by binding to RNA transcripts and blocking access from cellular machinery such as splicing factors. SBOs have the potential to bind to near-complementary sites in the transcriptome, causing off-target effects. In this study, we used RNA-seq to evaluate the off-target differential splicing events of 81 SBOs and differential expression events of 46 SBOs. Our results suggest that differential splicing events are predominantly hybridization driven, whereas differential expression events are more common and driven by other mechanisms (including spurious experimental variation). We further evaluated the performance of in silico screens for off-target splicing events, and found an edit distance cutoff of three to result in a sensitivity of 14% and false discovery rate (FDR) of 99%. A machine learning model incorporating splicing predictions substantially improved the ability to prioritize low edit distance hits, increasing sensitivity from 4% to 26% at a fixed FDR of 90%. Despite these large improvements in performance, this approach does not detect the majority of events at an FDR <99%. Our results suggest that in silico methods are currently of limited use for predicting the off-target effects of SBOs, and experimental screening by RNA-seq should be the preferred approach.

Deep learning in biomedicine.

Deep learning is beginning to impact biological research and biomedical applications as a result of its ability to integrate vast datasets, learn arbitrarily complex relationships and incorporate existing knowledge. Already, deep learning models can predict, with varying degrees of success, how genetic variation alters cellular processes involved in pathogenesis, which small molecules will modulate the activity of therapeutically relevant proteins, and whether radiographic images are indicative of disease. However, the flexibility of deep learning creates new challenges in guaranteeing the performance of deployed systems and in establishing trust with stakeholders, clinicians and regulators, who require a rationale for decision making. We argue that these challenges will be overcome using the same flexibility that created them; for example, by training deep models so that they can output a rationale for their predictions. Significant research in this direction will be needed to realize the full potential of deep learning in biomedicine.

Inference of the human polyadenylation code.

Motivation: Processing of transcripts at the 3'-end involves cleavage at a polyadenylation site followed by the addition of a poly(A)-tail. By selecting which site is cleaved, the process of alternative polyadenylation enables genes to produce transcript isoforms with different 3'-ends. To facilitate the identification and treatment of disease-causing mutations that affect polyadenylation and to understand the sequence determinants underlying this regulatory process, a computational model that can accurately predict polyadenylation patterns from genomic features is desirable. Results: Previous works have focused on identifying candidate polyadenylation sites and classifying tissue-specific sites. By training on how multiple sites in genes are competitively selected for polyadenylation from 3'-end sequencing data, we developed a deep learning model that can predict the tissue-specific strength of a polyadenylation site in the 3' untranslated region of the human genome given only its genomic sequence. We demonstrate the model's broad utility on multiple tasks, without any application-specific training. The model can be used to predict which polyadenylation site is more likely to be selected in genes with multiple sites. It can be used to scan the 3' untranslated region to find candidate polyadenylation sites. It can be used to classify the pathogenicity of variants near annotated polyadenylation sites in ClinVar. It can also be used to anticipate the effect of antisense oligonucleotide experiments to redirect polyadenylation. We provide analysis on how different features affect the model's predictive performance and a method to identify sensitive regions of the genome at the single-based resolution that can affect polyadenylation regulation. Supplementary information: Supplementary data are available at Bioinformatics online.

Local Submodularization for Binary Pairwise Energies.

Many computer vision problems require optimization of binary non-submodular energies. We propose a general optimization framework based on local submodular approximations (LSA). Unlike standard LP relaxation methods that linearize the whole energy globally, our approach iteratively approximates the energy locally. On the other hand, unlike standard local optimization methods (e.g., gradient descent or projection techniques) we use non-linear submodular approximations and optimize them without leaving the domain of integer solutions. We discuss two specific LSA algorithms based on trust region and auxiliary function principles, LSA-TR and LSA-AUX. The proposed methods obtain state-of-the-art results on a wide range of applications such as binary deconvolution, curvature regularization, inpainting, segmentation with repulsion and two types of shape priors. Finally, we discuss a move-making extension to the LSA-TR approach. While our paper is focused on pairwise energies, our ideas extend to higher-order problems. The code is available online.

Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning.

Knowing the sequence specificities of DNA- and RNA-binding proteins is essential for developing models of the regulatory processes in biological systems and for identifying causal disease variants. Here we show that sequence specificities can be ascertained from experimental data with 'deep learning' techniques, which offer a scalable, flexible and unified computational approach for pattern discovery. Using a diverse array of experimental data and evaluation metrics, we find that deep learning outperforms other state-of-the-art methods, even when training on in vitro data and testing on in vivo data. We call this approach DeepBind and have built a stand-alone software tool that is fully automatic and handles millions of sequences per experiment. Specificities determined by DeepBind are readily visualized as a weighted ensemble of position weight matrices or as a 'mutation map' that indicates how variations affect binding within a specific sequence.

The effects of amplitude normalization and EMG targets on cVEMP interaural amplitude asymmetry.

OBJECTIVES: The purpose of this investigation was to determine whether the use of visual feedback of tonic electromyographic (EMG) activity, or the use of amplitude normalization techniques would reduce significantly the variability in cervical vestibular evoked myogenic potential (cVEMP) P13-N23 interaural amplitude asymmetry data in otologically and neurologically intact children and adults. DESIGN: There were 97 subjects, both pediatric and adult, from whom the authors recorded cVEMPs with and without the use of an EMG target and amplitude normalization techniques. The four conditions were: (1) conventional recording (condition 1), (2) conventional recording with an EMG target (condition 2), (3) same as condition 1, with the addition of postacquisition amplitude normalization techniques, and (4) same as condition 2, with the addition of postacquisition amplitude normalization techniques. The absolute peak to peak amplitude of positive-negative response (P13-N23), absolute latency of P13, and the left or right amplitude asymmetry of P13-N23 were measured. RESULTS: Neither P13-N23 amplitudes nor latencies, neither mean root mean square (RMS) of the full wave rectified EMG activity, nor the standard deviation of the RMS EMG activity differed significantly when subjects were, and were not, asked to ensure their tonic EMG activity exceeded a visual target (i.e., representing >50 µV RMS EMG). Amplitude normalization of the cVEMP waveforms failed to reduce significantly the variability in the amplitude asymmetry data. CONCLUSIONS: Activating the sternocleidomastoid muscle with the patient in a semirecumbent position, with head turned away from the stimulated ear and head elevated (i.e., an optimal activation technique) was sufficient to produce the highest amplitude cVEMPs with an acceptable amount of variability in subjects of all ages. Group data suggested that the use of visual targets and amplitude normalization routines did not reduce significantly the variability in cVEMP interaural amplitude asymmetry measures. However, in isolated cases amplitude normalization converted an "abnormal" cVEMP into a "normal" cVEMP although the reverse occurred as well, suggesting that these techniques may be beneficial for a subset of patients receiving a less than perfectly administered test procedure.

Interactive Segmentation with Super-Labels.

In interactive segmentation, the most common way to model object appearance is by GMM or histogram, while MRFs are used to encourage spatial coherence among the object labels. This makes the strong assumption that pixels within each object are i.i.d. when in fact most objects have multiple distinct appearances and exhibit strong spatial correlation among their pixels. At the very least, this calls for an MRF-based appearance model within each object itself and yet, to the best of our knowledge, such a "two-level MRF" has never been proposed. We propose a novel segmentation energy that can model complex appearance. We represent the appearance of each object by a set of distinct spatially coherent models. This results in a two-level MRF with "super-labels" at the top level that are partitioned into "sub-labels" at the bottom. We introduce the hierarchical Potts (hPotts) prior to govern spatial coherence within each level. Finally, we introduce a novel algorithm with EM-style alternation of proposal, α-expansion and re-estimation steps. Our experiments demonstrate the conceptual and qualitative improvement that a two-level MRF can provide. We show applications in binary segmentation, multi-class segmentation, and interactive co-segmentation. Finally, our energy and algorithm have interesting interpretations in terms of semi-supervised learning.