FastPval

A fast and memory efficient program to calculate very low P-values from empirical distribution

Mulin Jun Li, Pak Chung Sham, Junwen Wang

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

Motivation: Resampling methods, such as permutation and bootstrap, have been widely used to generate an empirical distribution for assessing the statistical significance of a measurement. However, to obtain a very low P-value, a large size of resampling is required, where computing speed, memory and storage consumption become bottlenecks, and sometimes become impossible, even on a computer cluster. Results: We have developed a multiple stage P-value calculating program called FastPval that can efficiently calculate very low (up to 10-9) P-values from a large number of resampled measurements. With only two input files and a few parameter settings from the users, the program can compute P-values from empirical distribution very efficiently, even on a personal computer. When tested on the order of 109 resampled data, our method only uses 52.94% the time used by the conventional method, implemented by standard quicksort and binary search algorithms, and consumes only 0.11% of the memory and storage. Furthermore, our method can be applied to extra large datasets that the conventional method fails to calculate. The accuracy of the method was tested on data generated from Normal, Poison and Gumbel distributions and was found to be no different from the exact ranking approach.

Original languageEnglish (US)
Article numberbtq540
Pages (from-to)2897-2899
Number of pages3
JournalBioinformatics
Volume26
Issue number22
DOIs
StatePublished - Nov 2010
Externally publishedYes

Fingerprint

Empirical Distribution
Data storage equipment
Calculate
Poisons
Personal computers
Quicksort
Gumbel Distribution
Binary search
Resampling Methods
Statistical Distributions
Statistical Significance
Personal Computer
Resampling
Large Data Sets
Bootstrap
Microcomputers
Search Algorithm
Gaussian distribution
Ranking
Permutation

ASJC Scopus subject areas

  • Biochemistry
  • Molecular Biology
  • Computational Theory and Mathematics
  • Computer Science Applications
  • Computational Mathematics
  • Statistics and Probability
  • Medicine(all)

Cite this

FastPval : A fast and memory efficient program to calculate very low P-values from empirical distribution. / Li, Mulin Jun; Sham, Pak Chung; Wang, Junwen.

In: Bioinformatics, Vol. 26, No. 22, btq540, 11.2010, p. 2897-2899.

Research output: Contribution to journalArticle

@article{13c5784fe96c48478a9e97784d1c603d,
title = "FastPval: A fast and memory efficient program to calculate very low P-values from empirical distribution",
abstract = "Motivation: Resampling methods, such as permutation and bootstrap, have been widely used to generate an empirical distribution for assessing the statistical significance of a measurement. However, to obtain a very low P-value, a large size of resampling is required, where computing speed, memory and storage consumption become bottlenecks, and sometimes become impossible, even on a computer cluster. Results: We have developed a multiple stage P-value calculating program called FastPval that can efficiently calculate very low (up to 10-9) P-values from a large number of resampled measurements. With only two input files and a few parameter settings from the users, the program can compute P-values from empirical distribution very efficiently, even on a personal computer. When tested on the order of 109 resampled data, our method only uses 52.94{\%} the time used by the conventional method, implemented by standard quicksort and binary search algorithms, and consumes only 0.11{\%} of the memory and storage. Furthermore, our method can be applied to extra large datasets that the conventional method fails to calculate. The accuracy of the method was tested on data generated from Normal, Poison and Gumbel distributions and was found to be no different from the exact ranking approach.",
author = "Li, {Mulin Jun} and Sham, {Pak Chung} and Junwen Wang",
year = "2010",
month = "11",
doi = "10.1093/bioinformatics/btq540",
language = "English (US)",
volume = "26",
pages = "2897--2899",
journal = "Bioinformatics",
issn = "1367-4803",
publisher = "Oxford University Press",
number = "22",

}

TY - JOUR

T1 - FastPval

T2 - A fast and memory efficient program to calculate very low P-values from empirical distribution

AU - Li, Mulin Jun

AU - Sham, Pak Chung

AU - Wang, Junwen

PY - 2010/11

Y1 - 2010/11

N2 - Motivation: Resampling methods, such as permutation and bootstrap, have been widely used to generate an empirical distribution for assessing the statistical significance of a measurement. However, to obtain a very low P-value, a large size of resampling is required, where computing speed, memory and storage consumption become bottlenecks, and sometimes become impossible, even on a computer cluster. Results: We have developed a multiple stage P-value calculating program called FastPval that can efficiently calculate very low (up to 10-9) P-values from a large number of resampled measurements. With only two input files and a few parameter settings from the users, the program can compute P-values from empirical distribution very efficiently, even on a personal computer. When tested on the order of 109 resampled data, our method only uses 52.94% the time used by the conventional method, implemented by standard quicksort and binary search algorithms, and consumes only 0.11% of the memory and storage. Furthermore, our method can be applied to extra large datasets that the conventional method fails to calculate. The accuracy of the method was tested on data generated from Normal, Poison and Gumbel distributions and was found to be no different from the exact ranking approach.

AB - Motivation: Resampling methods, such as permutation and bootstrap, have been widely used to generate an empirical distribution for assessing the statistical significance of a measurement. However, to obtain a very low P-value, a large size of resampling is required, where computing speed, memory and storage consumption become bottlenecks, and sometimes become impossible, even on a computer cluster. Results: We have developed a multiple stage P-value calculating program called FastPval that can efficiently calculate very low (up to 10-9) P-values from a large number of resampled measurements. With only two input files and a few parameter settings from the users, the program can compute P-values from empirical distribution very efficiently, even on a personal computer. When tested on the order of 109 resampled data, our method only uses 52.94% the time used by the conventional method, implemented by standard quicksort and binary search algorithms, and consumes only 0.11% of the memory and storage. Furthermore, our method can be applied to extra large datasets that the conventional method fails to calculate. The accuracy of the method was tested on data generated from Normal, Poison and Gumbel distributions and was found to be no different from the exact ranking approach.

UR - http://www.scopus.com/inward/record.url?scp=78149251209&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78149251209&partnerID=8YFLogxK

U2 - 10.1093/bioinformatics/btq540

DO - 10.1093/bioinformatics/btq540

M3 - Article

VL - 26

SP - 2897

EP - 2899

JO - Bioinformatics

JF - Bioinformatics

SN - 1367-4803

IS - 22

M1 - btq540

ER -