Large-scale design and refinement of stable proteins using sequence-only models.

Engineered proteins generally must possess a stable structure in order to achieve their designed function. Stable designs, however, are astronomically rare within the space of all possible amino acid sequences. As a consequence, many designs must be tested computationally and experimentally in order to find stable ones, which is expensive in terms of time and resources. Here we use a high-throughput, low-fidelity assay to experimentally evaluate the stability of approximately 200,000 novel proteins. These include a wide range of sequence perturbations, providing a baseline for future work in the field. We build a neural network model that predicts protein stability given only sequences of amino acids, and compare its performance to the assayed values. We also report another network model that is able to generate the amino acid sequences of novel stable proteins given requested secondary sequences. Finally, we show that the predictive model-despite weaknesses including a noisy data set-can be used to substantially increase the stability of both expert-designed and model-generated proteins.

Prediction of whole-cell transcriptional response with machine learning.

MOTIVATION: Applications in synthetic and systems biology can benefit from measuring whole-cell response to biochemical perturbations. Execution of experiments to cover all possible combinations of perturbations is infeasible. In this paper, we present the host response model (HRM), a machine learning approach that maps response of single perturbations to transcriptional response of the combination of perturbations. RESULTS: The HRM combines high-throughput sequencing with machine learning to infer links between experimental context, prior knowledge of cell regulatory networks, and RNASeq data to predict a gene's dysregulation. We find that the HRM can predict the directionality of dysregulation to a combination of inducers with an accuracy of >90% using data from single inducers. We further find that the use of prior, known cell regulatory networks doubles the predictive performance of the HRM (an R2 from 0.3 to 0.65). The model was validated in two organisms, Escherichia coli and Bacillus subtilis, using new experiments conducted after training. Finally, while the HRM is trained with gene expression data, the direct prediction of differential expression makes it possible to also conduct enrichment analyses using its predictions. We show that the HRM can accurately classify >95% of the pathway regulations. The HRM reduces the number of RNASeq experiments needed as responses can be tested in silico prior to the experiment. AVAILABILITY AND IMPLEMENTATION: The HRM software and tutorial are available at https://github.com/sd2e/CDM and the configurable differential expression analysis tools and tutorials are available at https://github.com/SD2E/omics_tools. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Improved datasets and evaluation methods for the automatic prediction of DNA-binding proteins.

MOTIVATION: Accurate automatic annotation of protein function relies on both innovative models and robust datasets. Due to their importance in biological processes, the identification of DNA-binding proteins directly from protein sequence has been the focus of many studies. However, the datasets used to train and evaluate these methods have suffered from substantial flaws. We describe some of the weaknesses of the datasets used in previous DNA-binding protein literature and provide several new datasets addressing these problems. We suggest new evaluative benchmark tasks that more realistically assess real-world performance for protein annotation models. We propose a simple new model for the prediction of DNA-binding proteins and compare its performance on the improved datasets to two previously published models. In addition, we provide extensive tests showing how the best models predict across taxa. RESULTS: Our new gradient boosting model, which uses features derived from a published protein language model, outperforms the earlier models. Perhaps surprisingly, so does a baseline nearest neighbor model using BLAST percent identity. We evaluate the sensitivity of these models to perturbations of DNA-binding regions and control regions of protein sequences. The successful data-driven models learn to focus on DNA-binding regions. When predicting across taxa, the best models are highly accurate across species in the same kingdom and can provide some information when predicting across kingdoms. AVAILABILITY AND IMPLEMENTATION: The data and results for this article can be found at https://doi.org/10.5281/zenodo.5153906. The code for this article can be found at https://doi.org/10.5281/zenodo.5153683. The code, data and results can also be found at https://github.com/AZaitzeff/tools_for_dna_binding_proteins.

The Role of Configurational Entropy in Miniprotein Stability.

Predicting protein stability is a challenge due to the many competing thermodynamic effects. Through de novo protein design, one begins with a target structure and searches for a sequence that will fold into it. Previous work by Rocklin et al. introduced a data set of more than 16,000 miniproteins spanning four structural topologies with information on stability. These structures were characterized with a set of 46 structural descriptors, with no explicit inclusion of configurational entropy (Scnf). Our work focused on creating a set of 17 descriptors intended to capture variations in Scnf and its comparison to an extended set of 113 structural and energy model features that extend the Rocklin et al. feature set (R). The Scnf descriptors statistically discriminate between stable and unstable distributions within topologies and best describe EEHEE topology stability (where E = ß sheet and H = α helix). Between 50 and 80% of the variation in each Scnf descriptor is described by linear combinations of R features. Despite containing useful information about minipeptide stability, providing Scnf features as inputs to machine learning models does not improve overall performance when predicting protein stability, as the R features sufficiently capture the implicit variations.

What is changing when: Decoding visual information in movies from human intracranial recordings.

The majority of visual recognition studies have focused on the neural responses to repeated presentations of static stimuli with abrupt and well-defined onset and offset times. In contrast, natural vision involves unique renderings of visual inputs that are continuously changing without explicitly defined temporal transitions. Here we considered commercial movies as a coarse proxy to natural vision. We recorded intracranial field potential signals from 1,284 electrodes implanted in 15 patients with epilepsy while the subjects passively viewed commercial movies. We could rapidly detect large changes in the visual inputs within approximately 100 ms of their occurrence, using exclusively field potential signals from ventral visual cortical areas including the inferior temporal gyrus and inferior occipital gyrus. Furthermore, we could decode the content of those visual changes even in a single movie presentation, generalizing across the wide range of transformations present in a movie. These results present a methodological framework for studying cognition during dynamic and natural vision.

Sensitivity to timing and order in human visual cortex.

Visual recognition takes a small fraction of a second and relies on the cascade of signals along the ventral visual stream. Given the rapid path through multiple processing steps between photoreceptors and higher visual areas, information must progress from stage to stage very quickly. This rapid progression of information suggests that fine temporal details of the neural response may be important to the brain's encoding of visual signals. We investigated how changes in the relative timing of incoming visual stimulation affect the representation of object information by recording intracranial field potentials along the human ventral visual stream while subjects recognized objects whose parts were presented with varying asynchrony. Visual responses along the ventral stream were sensitive to timing differences as small as 17 ms between parts. In particular, there was a strong dependency on the temporal order of stimulus presentation, even at short asynchronies. From these observations we infer that the neural representation of complex information in visual cortex can be modulated by rapid dynamics on scales of tens of milliseconds.

Localization of sleep spindles, k-complexes, and vertex waves with subdural electrodes in children.

PURPOSE: To describe for the first time in children the localization of sleep spindles, K-complexes, and vertex waves using subdural electrodes. METHODS: We enrolled children who underwent presurgical evaluation of refractory epilepsy with subdural grid electrodes. We analyzed electroencephalogram data from subdural electrodes and simultaneous recording with Cz scalp electrode. Sleep spindles, K-complexes, and vertex waves were identified and localized based on their morphology on the subdural electrodes. RESULTS: Sixteen patients (9 boys; age range, 3-18 years) were enrolled in the study. The inter-rater reliability on identification and localization of maximal amplitude was high with an intraclass correlation coefficient of 0.85 for vertex waves, 0.94 for sleep spindles, and 0.91 for K-complexes. Sleep spindles presented maximum amplitude around the perirolandic area with a field extending to the frontal regions. K-complexes presented maximum amplitude around the perirolandic area with a field extending to the frontal regions. Vertex waves presented maximum amplitude around the perirolandic areas. CONCLUSIONS: In our series of pediatric patients, sleep spindles, K-complexes, and vertex waves were localized around the perirolandic area.

Short temporal asynchrony disrupts visual object recognition.

Humans can recognize objects and scenes in a small fraction of a second. The cascade of signals underlying rapid recognition might be disrupted by temporally jittering different parts of complex objects. Here we investigated the time course over which shape information can be integrated to allow for recognition of complex objects. We presented fragments of object images in an asynchronous fashion and behaviorally evaluated categorization performance. We observed that visual recognition was significantly disrupted by asynchronies of approximately 30 ms, suggesting that spatiotemporal integration begins to break down with even small deviations from simultaneity. However, moderate temporal asynchrony did not completely obliterate recognition; in fact, integration of visual shape information persisted even with an asynchrony of 100 ms. We describe the data with a concise model based on the dynamic reduction of uncertainty about what image was presented. These results emphasize the importance of timing in visual processing and provide strong constraints for the development of dynamical models of visual shape recognition.

Temporal stability of visually selective responses in intracranial field potentials recorded from human occipital and temporal lobes.

The cerebral cortex needs to maintain information for long time periods while at the same time being capable of learning and adapting to changes. The degree of stability of physiological signals in the human brain in response to external stimuli over temporal scales spanning hours to days remains unclear. Here, we quantitatively assessed the stability across sessions of visually selective intracranial field potentials (IFPs) elicited by brief flashes of visual stimuli presented to 27 subjects. The interval between sessions ranged from hours to multiple days. We considered electrodes that showed robust visual selectivity to different shapes; these electrodes were typically located in the inferior occipital gyrus, the inferior temporal cortex, and the fusiform gyrus. We found that IFP responses showed a strong degree of stability across sessions. This stability was evident in averaged responses as well as single-trial decoding analyses, at the image exemplar level as well as at the category level, across different parts of visual cortex, and for three different visual recognition tasks. These results establish a quantitative evaluation of the degree of stationarity of visually selective IFP responses within and across sessions and provide a baseline for studies of cortical plasticity and for the development of brain-machine interfaces.

Quickest detection of drug-resistant seizures: an optimal control approach.

Epilepsy affects 50 million people worldwide, and seizures in 30% of the cases remain drug resistant. This has increased interest in responsive neurostimulation, which is most effective when administered during seizure onset. We propose a novel framework for seizure onset detection that involves (i) constructing statistics from multichannel intracranial EEG (iEEG) to distinguish nonictal versus ictal states; (ii) modeling the dynamics of these statistics in each state and the state transitions; you can remove this word if there is no room. (iii) developing an optimal control-based "quickest detection" (QD) strategy to estimate the transition times from nonictal to ictal states from sequential iEEG measurements. The QD strategy minimizes a cost function of detection delay and false positive probability. The solution is a threshold that non-monotonically decreases over time and avoids responding to rare events that normally trigger false positives. We applied QD to four drug resistant epileptic patients (168 hour continuous recordings, 26-44 electrodes, 33 seizures) and achieved 100% sensitivity with low false positive rates (0.16 false positive/hour). This article is part of a Supplemental Special Issue entitled The Future of Automated Seizure Detection and Prediction.

Temporal cortex neurons encode articulated actions as slow sequences of integrated poses.

Form and motion processing pathways of the primate visual system are known to be interconnected, but there has been surprisingly little investigation of how they interact at the cellular level. Here we explore this issue with a series of three electrophysiology experiments designed to reveal the sources of action selectivity in monkey temporal cortex neurons. Monkeys discriminated between actions performed by complex, richly textured, rendered bipedal figures and hands. The firing patterns of neurons contained enough information to discriminate the identity of the character, the action performed, and the particular conjunction of action and character. This suggests convergence of motion and form information within single cells. Form and motion information in isolation were both sufficient to drive action discrimination within these neurons, but removing form information caused a greater disruption to the original response. Finally, we investigated the temporal window across which visual information is integrated into a single pose (or, equivalently, the timing with which poses are differentiated). Temporal cortex neurons under normal conditions represent actions as sequences of poses integrated over approximately 120 ms. They receive both motion and form information, however, and can use either if the other is absent.

Toward unmasking the dynamics of visual perception.

Fisch et al. report in this issue of Neuron the results of an investigation of the neural correlates of conscious perception. They find an early, dramatic, and long-lasting gamma response in high-level visual areas, when (and only when) a rapidly presented image is perceived.

A method for the real-time rendering of formless dot field structure-from-motion stimuli.

The perception of visual motion relies on different computations and different neural substrates than the perception of static form. It is therefore useful to have psychophysical stimuli that carry mostly or entirely motion information, conveying little or nothing about form in any single frame. Structure-from-motion stimuli can sometimes achieve this dissociation, with some examples in studies of biological motion using point-light walkers. It is, however, generally not trivial to provide motion information without also providing static form information. The problem becomes more computationally difficult when the structures and the motions in question are complex. Here we present a technique by which an animated three-dimensional scene can be rendered in real-time as a pattern of dots. Each dot follows the trajectory of the underlying object in the animation, but each static frame of the animation appears to be a uniform random field of dots. The resulting stimuli capture motion vectors across arbitrary complex scenes, while providing virtually no instantaneous information about the structure of that scene. We also present the results of a psychophysical experiment demonstrating the efficacy and the limitations of the technique. The ability to create such stimuli on the fly allows for interactive adjustment and control of the stimuli, real-time parametric variations of structure and motion, and the creation of large libraries of actions without the need to pre-render a prohibitive number of movies. This technique provides a powerful tool for the dissociation of complex motion from static form.

Effects of long-term object familiarity on event-related potentials in the monkey.

Although some change in the neural representation of an object must occur as it becomes familiar, the nature of this change is not fully understood. In humans, it has been shown that the N170-an evoked visual potential-is enhanced for classes of objects for which people have visual expertise. In this study, we explored whether monkeys show a similar modulation in event-related potential (ERP) amplitude as a result of long-term familiarity by recording ERPs with chronically implanted electrodes over extended training periods spanning many sessions. In each of 3 experiments, we found larger amplitude visual evoked responses to highly familiar images for the time period of 120-250 ms after stimulus onset. This difference was found when the monkeys were trained in an individual-level discrimination task, in a task that required only color discrimination, and even following a viewing-only task. We thus observed this familiarity effect across several tasks and different object categories and further found that the difference between "familiar" and "novel" became smaller as the animals gained experience with the previously unfamiliar objects across multiple test sessions. These data suggest that changes in visual responses associated with familiarity are evident early in the evoked visual response, are robust, and may be automatic, driven at least in part by repeated object exposure.

XOR style tasks for testing visual object processing in monkeys.

Using visually complex stimuli, three monkeys learned visual exclusive-or (XOR) tasks that required detecting two way visual feature conjunctions. Monkeys with passive exposure to the test images, or prior experience, were quicker to acquire an XOR style task. Training on each pairwise comparison of the stimuli to be used in an XOR task provided nearly complete transfer when stimuli became intermingled in the full XOR task. Task mastery took longer, accuracy was lower, and response times were slower for conjunction stimuli. Rotating features of the XOR stimuli did not adversely effect recognition speed or accuracy.