Statistical inference of protein structural alignments using information and compression

Collier, James H.; Allison, Lloyd; Lesk, Arthur M.; Stuckey, Peter J.; Banda, Maria García de la; Konagurthu, Arun S.

doi:10.1101/056598

Cited by 2 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 1 shows the reference-dependent alignment accuracy on Test5.6K dataset. To reduce the potential bias in generating reference alignments, we evaluated our approach using reference alignments generated using three structure alignment tools, including TMalign [32], DeepAlign and MM-Ligner [6].…”

Section: Evaluation Methodsmentioning

confidence: 99%

ProALIGN: Directly learning alignments for protein structure prediction via exploiting context-specific alignment motifs

Kong

Zheng

et al. 2020

Preprint

View full text Add to dashboard Cite

Template-based modeling (TBM), including homology modeling and protein threading, is one of the most reliable techniques for protein structure prediction. It predicts protein structure by building an alignment between the query sequence under prediction and the templates with solved structures. However, it is still very challenging to build the optimal sequence-template alignment, especially when only distantly-related templates are available. Here we report a novel deep learning approach ProALIGN that can predict much more accurate sequence-template alignment. Like protein sequences consisting of sequence motifs, protein alignments are also composed of frequently-occurring alignment motifs with characteristic patterns. Alignment motifs are context-specific as their characteristic patterns are tightly related to sequence contexts of the aligned regions. Inspired by this observation, we represent a protein alignment as a binary matrix (in which 1 denotes an aligned residue pair) and then use a deep convolutional neural network to predict the optimal alignment from the query protein and its template. The trained neural network implicitly but effectively encodes an alignment scoring function, which reduces inaccuracies in the handcrafted scoring functions widely used by the current threading approaches. For a query protein and a template, we apply the neural network to directly infer likelihoods of all possible residue pairs in their entirety, which could effectively consider the correlations among multiple residues. We further construct the alignment with maximum likelihood, and finally build structure model according to the alignment.Tested on three independent datasets with in total 6,688 protein alignment targets and 80 CASP13 TBM targets, our method achieved much better alignments and 3D structure models than the existing methods including HHpred, CNFpred, CEthreader and DeepThreader. These results clearly demonstrate the effectiveness of exploiting the context-specific alignment motifs by deep learning for protein threading.

show abstract

Section: Evaluation Methodsmentioning

confidence: 99%

ProALIGN: Directly learning alignments for protein structure prediction via exploiting context-specific alignment motifs

Kong

Zheng

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…The above described 𝐶𝑜𝑠𝑡 )+2=> (𝑖, 𝑗) distance measure is computed using the standard MML Wallace Freeman approximation 21,54 defined for a Gaussian distribution 22,55,56 . As defined by Equation 1, for any dataset 𝐷 and a hypothesis 𝐻 that describes 𝐷 = [𝑥 !…”

Section: Computing the Total Encoding Message Length For Any Gaussian...mentioning

confidence: 99%

Gene-level alignment of single cell trajectories

Sumanaweera¹,

Suo²,

Muraro³

et al. 2023

Preprint

View full text Add to dashboard Cite

Single cell data analysis can infer dynamic changes in cell populations, for example across time, space or in response to perturbation. To compare these dynamics between two conditions, trajectory alignment via dynamic programming (DP) optimization is frequently used, but is limited by assumptions such as a definite existence of a match. Here we describe Genes2Genes, a Bayesian information-theoretic DP framework for aligning single-cell trajectories. Genes2Genes overcomes current limitations and is able to capture sequential matches and mismatches between a reference and a query at single gene resolution, highlighting distinct clusters of genes with varying patterns of gene expression dynamics. Across both real life and simulated datasets, Genes2Genes accurately captured different alignment patterns, and revealed that T cells differentiated in vitro matched to an immature in vivo state while lacking the final TNFɑ signaling. This use case demonstrates that precise trajectory alignment can pinpoint divergence from the in vivo system, thus providing an opportunity to optimize in vitro culture conditions.

show abstract

Statistical inference of protein structural alignments using information and compression

Cited by 2 publications

References 27 publications

ProALIGN: Directly learning alignments for protein structure prediction via exploiting context-specific alignment motifs

ProALIGN: Directly learning alignments for protein structure prediction via exploiting context-specific alignment motifs

Gene-level alignment of single cell trajectories

Contact Info

Product

Resources

About