Direct AUC optimization of regulatory motifs

Zhu, Lin; Zhang, Hongbo; Huang, De-Shuang

doi:10.1093/bioinformatics/btx255

Cited by 32 publications

(7 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the same time, we plan to use the CNV detection results in the correction of Cancer Cell Fraction (CCF), which will greatly promote a comprehensive understanding of tumor occurrence and development ( Xi et al, 2018 ; Mao et al, 2021 ; Tarabichi et al, 2021 ). We will also try to apply this algorithm to the research of ancient DNA mutation detection, which may be helpful to explore the evolution process of species ( Zhu et al, 2017b ).…”

Section: Discussionmentioning

confidence: 99%

HBOS-CNV: A New Approach to Detect Copy Number Variations From Next-Generation Sequencing Data

2021

View full text Add to dashboard Cite

Copy number variation (CNV) is a genomic mutation that plays an important role in tumor evolution and tumor genesis. Accurate detection of CNVs from next-generation sequencing (NGS) data is still a challenging task due to artifacts such as uneven mapped reads and unbalanced amplitudes of gains and losses. This study proposes a new approach called HBOS-CNV to detect CNVs from NGS data. The central point of HBOS-CNV is that it uses a new statistic, the histogram-based outlier score (HBOS), to evaluate the fluctuation of genome bins to determine those of changed copy numbers. In comparison with existing statistics in the evaluation of CNVs, HBOS is a non-linearly transformed value from the observed read depth (RD) value of each genome bin, having the potential ability to relieve the effects resulted from the above artifacts. In the calculation of HBOS values, a dynamic width histogram is utilized to depict the density of bins on the genome being analyzed, which can reduce the effects of noises partially contributed by mapping and sequencing errors. The evaluation of genome bins using such a new statistic can lead to less extremely significant CNVs having a high probability of detection. We evaluated this method using a large number of simulation datasets and compared it with four existing methods (CNVnator, CNV-IFTV, CNV-LOF, and iCopyDav). The results demonstrated that our proposed method outperforms the others in terms of sensitivity, precision, and F1-measure. Furthermore, we applied the proposed method to a set of real sequencing samples from the 1000 Genomes Project and determined a number of CNVs with biological meanings. Thus, the proposed method can be regarded as a routine approach in the field of genome mutation analysis for cancer samples.

show abstract

Section: Discussionmentioning

confidence: 99%

HBOS-CNV: A New Approach to Detect Copy Number Variations From Next-Generation Sequencing Data

2021

View full text Add to dashboard Cite

show abstract

“…One of the most popular genetic feature inputted for study analysis are SNPs (Single Nucleotide Polymorphisms), these are variants in base pairs within the DNA sequence [8] [6]. While a majority of these SNPs will have little to no impact on the biological systems, the consequential causal sequence can lead to imbalances in chemicals, misfolds in protein polypeptide chains and instability in mRNA transcripts [9] [11][12][13][14][15]. The involvement of these SNPs in the genetic analysis for the purpose of finding risk variants is due to the abundance of variation throughout the genome; proving promising and successful in many determined diseases so far [9][10] [16].…”

Section: Background and Related Workmentioning

confidence: 99%

A Novel Approach to Detecting Epistasis using Random Sampling Regularisation

Hind¹,

Lisboa

Hussain

et al. 2020

IEEE/ACM Trans. Comput. Biol. and Bioinf.

View full text Add to dashboard Cite

Epistasis is a progressive approach that complements the 'common disease, common variant' hypothesis that highlights the potential for connected networks of genetic variants collaborating to produce a phenotypic expression. Epistasis is commonly performed as a pairwise or limitless-arity capacity that considers variant networks as either variant vs variant or as high order interactions. This type of analysis extends the number of tests that were previously performed in a standard approach such as Genome-Wide Association Study (GWAS), in which False Discovery Rate (FDR) is already an issue, therefore by multiplying the number of tests up to a factorial rate also increases the issue of FDR. Further to this, epistasis introduces its own limitations of computational complexity and intensity that are generated based on the analysis performed; to consider the most intense approach, a multivariate analysis introduces a time complexity of O(n!). Proposed in this paper is a novel methodology for the detection of epistasis using interpretable methods and best practice to outline interactions through filtering processes. Using a process of Random Sampling Regularisation which randomly splits and produces sample sets to conduct a voting system to regularise the significance and reliability of biological markers, SNPs. Preliminary results are promising, outlining a concise detection of interactions. Results for the detection of epistasis, in the classification of breast cancer patients, indicated eight outlined risk candidate interactions from five variants and a singular candidate variant with high protective association.

show abstract

“…Direct optimization of AUC without gradient-based methods Optimizing AUC directly using coordinate descent (i.e. optimizing one model parameter at a time, which is feasible despite the zero gradient) yields good results for certain machine learning methods that were designed specifically for genetics applications [Zhu et al, 2017]. LeDell et al [2016] introduce an ensemble approach based on Super Learner [van der Laan et al, 2007] which adjusts the combination of scores from several individual classifiers in favor of a higher AUC.…”

Section: Difficulties Of Auc Optimization and Related Workmentioning

confidence: 99%

Deep Learning for Virtual Screening: Five Reasons to Use ROC Cost Functions

Golkov

Becker

Plop

et al. 2020

Preprint

View full text Add to dashboard Cite

Computer-aided drug discovery is an essential component of modern drug development. Therein, deep learning has become an important tool for rapid screening of billions of molecules in silico for potential hits containing desired chemical features. Despite its importance, substantial challenges persist in training these models, such as severe class imbalance, high decision thresholds, and lack of ground truth labels in some datasets. In this work we argue in favor of directly optimizing the receiver operating characteristic (ROC) in such cases, due to its robustness to class imbalance, its ability to compromise over different decision thresholds, certain freedom to influence the relative weights in this compromise, fidelity to typical benchmarking measures, and equivalence to positive/unlabeled learning. We also propose new training schemes (coherent mini-batch arrangement, and usage of out-of-batch samples) for cost functions based on the ROC, as well as a cost function based on the logAUC metric that facilitates early enrichment (i.e. improves performance at high decision thresholds, as often desired when synthesizing predicted hit compounds). We demonstrate that these approaches outperform standard deep learning approaches on a series of PubChem high-throughput screening datasets that represent realistic and diverse drug discovery campaigns on major drug target families.

show abstract

Direct AUC optimization of regulatory motifs

Cited by 32 publications

References 39 publications

HBOS-CNV: A New Approach to Detect Copy Number Variations From Next-Generation Sequencing Data

HBOS-CNV: A New Approach to Detect Copy Number Variations From Next-Generation Sequencing Data

A Novel Approach to Detecting Epistasis using Random Sampling Regularisation

Deep Learning for Virtual Screening: Five Reasons to Use ROC Cost Functions

Contact Info

Product

Resources

About