Marvel: A Data-Centric Approach for Mapping Deep Learning Operators on Spatial Accelerators

A spatial accelerator’s efficiency depends heavily on both its mapper and cost models to generate optimized mappings for various operators of DNN models. However, existing cost models lack a formal boundary over their input programs (operators) for accurate and tractable cost analysis of the mappings, and this results in adaptability challenges to the cost models for new operators. We consider the recently introduced Maestro Data-Centric (MDC) notation and its analytical cost model to address this challenge because any mapping expressed in the notation is precisely analyzable using the MDC’s cost model. In this article, we characterize the set of input operators and their mappings expressed in the MDC notation by introducing a set of conformability rules . The outcome of these rules is that any loop nest that is perfectly nested with affine tensor subscripts and without conditionals is conformable to the MDC notation. A majority of the primitive operators in deep learning are such loop nests. In addition, our rules enable us to automatically translate a mapping expressed in the loop nest form to MDC notation and use the MDC’s cost model to guide upstream mappers. Our conformability rules over the input operators result in a structured mapping space of the operators, which enables us to introduce a mapper based on our decoupled off-chip/on-chip approach to accelerate mapping space exploration. Our mapper decomposes the original higher-dimensional mapping space of operators into two lower-dimensional off-chip and on-chip subspaces and then optimizes the off-chip subspace followed by the on-chip subspace. We implemented our overall approach in a tool called Marvel , and a benefit of our approach is that it applies to any operator conformable with the MDC notation. We evaluated Marvel over major DNN operators and compared it with past optimizers.

A journey through mapping space: characterising the statistical and metric properties of reduced representations of macromolecules

A mapping of a macromolecule is a prescription to construct a simplified representation of the system in which only a subset of its constituent atoms is retained. As the specific choice of the mapping affects the analysis of all-atom simulations as well as the construction of coarse-grained models, the characterisation of the mapping space has recently attracted increasing attention. We here introduce a notion of scalar product and distance between reduced representations, which allows the study of the metric and topological properties of their space in a quantitative manner. Making use of a Wang–Landau enhanced sampling algorithm, we exhaustively explore such space, and examine the qualitative features of mappings in terms of their squared norm. A one-to-one correspondence with an interacting lattice gas on a finite volume leads to the emergence of discontinuous phase transitions in mapping space, which mark the boundaries between qualitatively different reduced representations of the same molecule. Graphicabstract

Mapping, Registering, and Ordering

This chapter examines imperial initiatives in mapping space, registering people, and ordering knowledge. The author draws a distinction between the mapping practices of pre-modern or tributary empires and those of early modern and modern imperial formations. In the latter case, authority was increasingly derived from the production and accumulation of knowledge via scientific techniques that relied on abstraction and quantification, whether at home or abroad. The author shows that modern imperial practices based on measurement were not limited to the West, but were also employed in the Ottoman Empire, Qing China, and parts of Mughal India. The chapter’s focus is the emergence of coordinate mapping as a tool of imperial expansion and control from the Renaissance through the mid-twentieth century. Similar techniques of legibility and quantification were applied to registering people and ordering knowledge. James C. Scott’s work on legibility in modern state building is foundational to this chapter.

DLDTI: a learning-based framework for drug-target interaction identification using neural networks and network representation

Background Drug repositioning, the strategy of unveiling novel targets of existing drugs could reduce costs and accelerate the pace of drug development. To elucidate the novel molecular mechanism of known drugs, considering the long time and high cost of experimental determination, the efficient and feasible computational methods to predict the potential associations between drugs and targets are of great aid. Methods A novel calculation model for drug-target interaction (DTI) prediction based on network representation learning and convolutional neural networks, called DLDTI, was generated. The proposed approach simultaneously fused the topology of complex networks and diverse information from heterogeneous data sources, and coped with the noisy, incomplete, and high-dimensional nature of large-scale biological data by learning the low-dimensional and rich depth features of drugs and proteins. The low-dimensional feature vectors were used to train DLDTI to obtain the optimal mapping space and to infer new DTIs by ranking candidates according to their proximity to the optimal mapping space. More specifically, based on the results from the DLDTI, we experimentally validated the predicted targets of tetramethylpyrazine (TMPZ) on atherosclerosis progression in vivo. Results The experimental results showed that the DLDTI model achieved promising performance under fivefold cross-validations with AUC values of 0.9172, which was higher than the methods using different classifiers or different feature combination methods mentioned in this paper. For the validation study of TMPZ on atherosclerosis, a total of 288 targets were identified and 190 of them were involved in platelet activation. The pathway analysis indicated signaling pathways, namely PI3K/Akt, cAMP and calcium pathways might be the potential targets. Effects and molecular mechanism of TMPZ on atherosclerosis were experimentally confirmed in animal models. Conclusions DLDTI model can serve as a useful tool to provide promising DTI candidates for experimental validation. Based on the predicted results of DLDTI model, we found TMPZ could attenuate atherosclerosis by inhibiting signal transductions in platelets. The source code and datasets explored in this work are available at https://github.com/CUMTzackGit/DLDTI.

DLDTI: A learning-based framework for drug-target interaction identification using neural networks and network representation

Background: Drug repositioning, the strategy of unveiling novel targets of existing drugs could reduce costs and accelerate the pace of drug development. To elucidate the novel molecular mechanism of known drugs, considering the long time and high cost of experimental determination, the efficient and feasible computational methods to predict the potential associations between drugs and targets are of great aid.Methods: A novel calculation model for drug-target interaction (DTI) prediction based on network representation learning and convolutional neural networks, called DLDTI， was generated. The proposed approach simultaneously fuses the topology of complex networks and diverse information from heterogeneous data sources, and copes with the noisy, incomplete, and high-dimensional nature of large-scale biological data by learning the low-dimensional and rich depth features of drugs and proteins. The low-dimensional feature vectors were used to train DLDTI to obtain the optimal mapping space and to infer new DTIs by ranking candidates according to their proximity to the optimal mapping space. More specifically, based on the results from the DLDTI, we experimentally validate the predicted targets of tetramethylpyrazine (TMPZ) on atherosclerosis progression in vivo.Results: The experimental results show that the DLDTI model achieves promising performance under 5-fold cross-validations with AUC values of 0.9172, which is higher than the methods using different classifiers or different feature combination methods mentioned in this paper. For the validation study of TMPZ on atherosclerosis, a total of 288 targets were identified and 190 of them were involved in platelet activation. The pathway analysis indicated signaling pathways, namely PI3K/Akt, cAMP and calcium pathways might be the potential targets. Effects and molecular mechanism of TMPZ on atherosclerosis were experimentally confirmed in animal models.Conclusions: DLDTI model can serve as a useful tool to provide promising DTI candidates for experimental validation. Based on the predicted results of DLDTI model, we found TMPZ could attenuate atherosclerosis by inhibiting signal transductions in platelets. The source code and datasets explored in this work are available at https://github.com/CUMTzackGit/DLDTI.