kNN-Borůvka-GPU: A Fast and Scalable MST Construction from kNN Graphs on GPU

Arefin, Ahmed Shamsul; Riveros, Carlos; Berretta, Regina; Moscato, Pablo

doi:10.1007/978-3-642-31125-3_6

Cited by 10 publications

(2 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We modify NVIDIA’s [35] HDBSCAN implementation for euclidean minimum spanning tree construction which finds the k -nearest nearbors using FAISS’s k -select [36] fused with a matrix multiply kernel to calculate the euclidean k -nearest neighbors, thereafter finding the Euclidean-Minimum Spanning Tree [37].…”

Section: Methodsmentioning

confidence: 99%

GenomeFace: a deep learning-based metagenome binner trained on 43,000 microbial genomes

Lettich,

Egan,

Riley

et al. 2024

Preprint

View full text Add to dashboard Cite

Metagenomic binning, the process of grouping DNA sequences into taxonomic units, is critical for understanding the functions, interactions, and evolutionary dynamics of microbial communities. We propose a deep learning approach to binning using two neural networks, one based on composition and another on environmental abundance, dynamically weighting the contribution of each based on characteristics of the input data. Trained on over 43,000 prokaryotic genomes, our network for composition-based binning is inspired by metric learning techniques used for facial recognition. Using a task-specific, multi-GPU accelerated algorithm to cluster the embeddings produced by our network, our binner leverages marker genes observed to be universally present in nearly all taxa to grade and select optimal clusters of sequences from a hierarchy of candidates. We evaluate our approach on four simulated datasets with known ground truth. Our linear time integration of marker genes recovers more near complete genomes than state of the art but computationally infeasible solutions using them, while being over an order of magnitude faster. Finally, we demonstrate the scalability and acuity of our approach by testing it on three of the largest metagenome assemblies ever performed. Compared to other binners, we produced 47%-183% more near complete genomes. From these datasets, we find over the genomes of over 3000 new candidate species which have never been previously cataloged, representing a potential 4% expansion of the known bacterial tree of life.

show abstract

Section: Methodsmentioning

confidence: 99%

GenomeFace: a deep learning-based metagenome binner trained on 43,000 microbial genomes

Lettich,

Egan,

Riley

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Nasre et al [84] proposed efficient techniques to execute subgraph addition, deletion, conflict detection and some optimization techniques to improve the performance of the MST algorithm such as employing an adaptive scheme to flexibly change kernel configurations. Arefin et al [85] solved the MST problem on GPUs by proposing a solution which combines the classical Boruvka's algorithm and the k nearest neighbor (kNN) graph data structure. The method first generated a kNN graph based on the original graph.…”

Section: Single-sourcementioning

confidence: 99%

GPU-based commonsense reasoning for real-time query answering and multimodal analysis

Tran¹

View full text Add to dashboard Cite

am truly privileged to have learned from his deep knowledge and research enthusiasm. Since the beginning of my research, he has inspired me and given me a passion for research. He has always shown me great research directions and unique ideas, along with encouragement and unwavering support regardless of his time. As well, I would greatly thank Prof. Kim Jung-Jae, Prof.He Bingsheng, Prof. James Cheng, and Prof. Foh Chuan Heng, for their helpful comments and advices.In addition, I also want to thank all my co-authors, my colleagues in Nanyang Technological University and my friends for their unconditional help. I gratefully acknowledge the scholarship towards my PhD from Nanyang Technological University.Last but not least, my deepest love and gratitude are devoted to all of my family members: my parents, my brothers, and especially my wife. Their loves will be endless motivation for me to strive and help me to overcome difficult times in my life.

show abstract