Ning Qian

Predicting the secondary structure of globular proteins using neural network models

We present a new method for predicting the secondary structure of globular proteins based on non-linear neural network models. Network models learn from existing protein structures how to predict the secondary structure of local sequences of amino acids. The average success rate of our method on a testing set of proteins non-homologous with the corresponding training set was 64.3% on three types of secondary structure (alpha-helix, beta-sheet, and coil), with correlation coefficients of C alpha = 0.41, C beta = 0.31 and Ccoil = 0.41. These quality indices are all higher than those of previous methods. The prediction accuracy for the first 25 residues of the N-terminal sequence was significantly better. We conclude from computational experiments on real and artificial structures that no method based solely on local information in the protein sequence is likely to produce significantly better results for non-homologous proteins. The performance of our method of homologous proteins is much better than for non-homologous proteins, but is not as good as simply assuming that homologous sequences have identical structures.

Practising orientation identification improves orientation coding in V1 neurons.

The adult brain shows remarkable plasticity, as demonstrated by the improvement in fine sensorial discriminations after intensive practice. The behavioural aspects of such perceptual learning are well documented, especially in the visual system. Specificity for stimulus attributes clearly implicates an early cortical site, where receptive fields retain fine selectivity for these attributes; however, the neuronal correlates of a simple visual discrimination task remained unidentified. Here we report electrophysiological correlates in the primary visual cortex (V1) of monkeys for learning orientation identification. We link the behavioural improvement in this type of learning to an improved neuronal performance of trained compared to naive neurons. Improved long-term neuronal performance resulted from changes in the characteristics of orientation tuning of individual neurons. More particularly, the slope of the orientation tuning curve that was measured at the trained orientation increased only for the subgroup of trained neurons most likely to code the orientation identified by the monkey. No modifications of the tuning curve were observed for orientations for which the monkey had not been trained. Thus training induces a specific and efficient increase in neuronal sensitivity in V1.

On the momentum term in gradient descent learning algorithms

A momentum term is usually included in the simulations of connectionist learning algorithms. Although it is well known that such a term greatly improves the speed of learning, there have been few rigorous studies of its mechanisms. In this paper, I show that in the limit of continuous time, the momentum parameter is analogous to the mass of Newtonian particles that move through a viscous medium in a conservative force field. The behavior of the system near a local minimum is equivalent to a set of coupled and damped harmonic oscillators. The momentum term improves the speed of convergence by bringing some eigen components of the system closer to critical damping. Similar results can be obtained for the discrete time case used in computer simulations. In particular, I derive the bounds for convergence on learning-rate and momentum parameters, and demonstrate that the momentum term can increase the range of learning rate over which the system converges. The optimal condition for convergence is also analyzed.

Computing stereo disparity and motion with known binocular cell properties

Many models for stereo disparity computation have been proposed, but few can be said to be truly biological. There is also a rich literature devoted to physiological studies of stereopsis. Cells sensitive to binocular disparity have been found in the visual cortex, but it is not clear whether these cells could be used to compute disparity maps from stereograms. Here we propose a model for biological stereo vision based on known receptive field profiles of binocular cells in the visual cortex and provide the first demonstration that these cells could effectively solve random dot stereograms. Our model also allows a natural integration of stereo vision and motion detection. This may help explain the existence of units tuned to both disparity and motion in the visual cortex.

Binocular Disparity and the Perception of Depth

I would like to thank my collaborators Drs. Yudong Zhu and Richard Andersen for their contributions to some of the works reviewed here. I am also grateful to Drs. Eric Kandel, Bard Geesaman, John Koester, and Vicent Ferrera for their very helpful discussions, comments, and suggestions on earlier versions of the manuscript. This review was supported by a research grant from the McDonnell-Pew Program in Cognitive Neuroscience and NIH grant Number MH54125.

Physiological computation of binocular disparity

We previously proposed a physiologically realistic model for stereo vision based on the quantitative binocular receptive field profiles mapped by Freeman and coworkers. Here we present several new results about the model that shed light on the physiological processes involved in disparity computation. First, we show that our model can be extended to a much more general class of receptive field profiles than the commonly used Gabor functions. Second, we demonstrate that there is, however, an advantage of using the Gabor filters: similar to our perception, the stereo algorithm with the Gabor filters has a small bias towards zero disparity. Third, we prove that the complex cells as described by Freeman et al. compute disparity by effectively summing up two related cross products between the band-pass filtered left and right retinal image patches. This operation is related to cross-correlation but it overcomes some major problems with the standard correlator. Fourth, we demonstrate that as few as two complex cells at each spatial location are sufficient for a reasonable estimation of binocular disparity. Fifth, we find that our model can be significantly improved by considering the fact that complex cell receptive field are, on average, larger than those of simple cells. This fact is incorporated into the model by averaging over several quadrature pairs of simple cells with nearby and overlapping receptive fields to construct a model complex cell. The disparity tuning curve of the resulting complex cell is much more reliable than the constructed from a single quadrature pair of simple cells used previously, and the computed disparity maps for random dot stereograms with the new algorithm are very similar to human perception, with sharp transitions at disparity boundaries. Finally, we show that under most circumstances our algorithm works equally well with either of the two well-known receptive field models in the literature.

An electro-diffusion model for computing membrane potentials and ionic concentrations in branching dendrites, spines and axons

The Nernst-Planck equation for electrodiffusion was applied to axons, dendrites and spines. For thick processes (1 μm) the results of computer simulation agreed accurately with the cable model for passive conduction and for propagating action potentials. For thin processes (0.1 μm) and spines, however, the cable model may fail during transient events such as synaptic potentials. First, ionic concentrations can rapidly change in small compartments, altering ionic equilibrium potentials and the driving forces for movement of ions across the membrane. Second, longitudinal diffusion may dominate over electrical forces when ionic concentration gradients become large. We compare predictions of the cable model and the electro-diffusion model for excitatory postsynaptic potentials on spines and show that there are significant discrepancies for large conductance changes. The electro-diffusion model also predicts that inhibition on small structures such as spines and thin processes is ineffective. We suggest a modified cable model that gives better agreement with the electro-diffusion model.

A physiological model for motion-stereo integration and a unified explanation of Pulfrich-like phenomena.

Many psychophysical and physiological experiments indicate that visual motion analysis and stereoscopic depth perception are processed together in the brain. However, little computational effort has been devoted to combining these two visual modalities into a common framework based on physiological mechanisms. We present such an integrated model in this paper. We have previously developed a physiologically realistic model for binocular disparity computation (Qian, 1994). Here we demonstrate that under some general and physiological assumptions, our stereo vision model can be combined naturally with motion energy models to achieve motion-stereo integration. The integrated model may be used to explain a wide range of experimental observations regarding motion-stereo interaction. As an example, we show that the model can provide a unified account of the classical Pulfrich effect (Morgan & Thompson, 1975) and the generalized Pulfrich phenomena to dynamic noise patterns (Tyler, 1974; Falk, 1980) and stroboscopic stimuli (Burr & Ross, 1979).

A coarse-to-fine disparity energy model with both phase-shift and position-shift receptive field mechanisms

Numerous studies suggest that the visual system uses both phase-and position-shift receptive field (RF) mechanisms for the processing of binocular disparity. Although the difference between these two mechanisms has been analyzed before, previous work mainly focused on disparity tuning curves instead of population responses. However, tuning curve and population response can exhibit different characteristics, and it is the latter that determines disparity estimation. Here we demonstrate, in the framework of the disparity energy model, that for relatively small disparities, the population response generated by the phase-shift mechanism is more reliable than that generated by the position-shift mechanism. This is true over a wide range of parameters, including the RF orientation. Since the phase model has its own drawbacks of underestimating large stimulus disparity and covering only a restricted range of disparity at a given scale, we propose a coarse-to-fine algorithm for disparity computation with a hybrid of phase-shift and position-shift components. In this algorithm, disparity at each scale is always estimated by the phase-shift mechanism to take advantage of its higher reliability. Since the phase-based estimation is most accurate at the smallest scale when the disparity is correspondingly small, the algorithm iteratively reduces the input disparity from coarse to fine scales by introducing a constant position-shift component to all cells for a given location in order to offset the stimulus disparity at that location. The model also incorporates orientation pooling and spatial pooling to further enhance reliability. We have tested the algorithm on both synthetic and natural stereo images and found that it often performs better than a simple scale-averaging procedure.

Adaptation across the Cortical Hierarchy: Low-Level Curve Adaptation Affects High-Level Facial-Expression Judgments

Adaptation is ubiquitous in sensory processing. Although sensory processing is hierarchical, with neurons at higher levels exhibiting greater degrees of tuning complexity and invariance than those at lower levels, few experimental or theoretical studies address how adaptation at one hierarchical level affects processing at others. Nevertheless, this issue is critical for understanding cortical coding and computation. Therefore, we examined whether perception of high-level facial expressions can be affected by adaptation to low-level curves (i.e., the shape of a mouth). After adapting to a concave curve, subjects more frequently perceived faces as happy, and after adapting to a convex curve, subjects more frequently perceived faces as sad. We observed this multilevel aftereffect with both cartoon and real test faces when the adapting curve and the mouths of the test faces had the same location. However, when we placed the adapting curve 0.2° below the test faces, the effect disappeared. Surprisingly, this positional specificity held even when real faces, instead of curves, were the adapting stimuli, suggesting that it is a general property for facial-expression aftereffects. We also studied the converse question of whether face adaptation affects curvature judgments, and found such effects after adapting to a cartoon face, but not a real face. Our results suggest that there is a local component in facial-expression representation, in addition to holistic representations emphasized in previous studies. By showing that adaptation can propagate up the cortical hierarchy, our findings also challenge existing functional accounts of adaptation.