Skip to main content

Table 1 Balanced Accuracy of optimal multivariable regression models

From: Feature-specific quantile normalization and feature-specific mean–variance normalization deliver robust bi-directional classification and feature selection performance between microarray and RNAseq data

 

Breast

Colon

Characteristic

Betaa

95% CIb

p-Value

Betaa

95% CIb

p-Value

Model

      

glmnet

–

–

 

–

–

 

SVM

 − 0.001

 − 0.009, 0.006

0.8

0.015

0.007, 0.023

 < 0.001

Train distribution

      

Agilent

–

–

 

–

–

 

RNAseq

 − 0.006

 − 0.015, 0.002

0.15

0.016

0.007, 0.026

 < 0.001

Test distribution

      

Agilent

–

–

 

–

–

 

RNAseq

0.006

 − 0.002, 0.015

0.15

0.034

0.025, 0.044

 < 0.001

Normalization method

      

Reference (REF)

–

–

 

–

–

 

FSQN

0.003

 − 0.008, 0.014

0.6

 − 0.006

 − 0.017, 0.005

0.3

FSMVN

 − 0.000

 − 0.011, 0.011

 > 0.9

 − 0.013

 − 0.024, − 0.001

0.031

LOG2

 − 0.347

 − 0.357, − 0.336

 < 0.001

 − 0.349

 − 0.360, − 0.337

 < 0.001

Feature selection method

      

Full Model

–

–

 

–

–

 

glmnet

0.009

0.001, 0.018

0.064

0.002

 − 0.008, 0.012

0.7

SVM

0.012

0.002, 0.021

0.015

0.009

0.001, 0.019

0.083

  1. aBeta = Percentage expressed as a decimal, bCI Confidence Interval
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us