Reproducibility of mass spectrometry based metabolomics data

BMC Bioinformatics

Table 2 Summary of reproducible metabolites per sample pair and reproducible sample pairs per metabolite for BioTech data set

Imputation	Normalization		Percentage of reproducible metabolites per sample pair			Percentage of reproducible sample pairs per metabolite
Imputation	Normalization		> 70%	> 80%	> 90%	> 70%	> 80%	> 90%
BPCA	Median	Biological layer	52.27	10.68	0.75	64.69	61.22	55.56
	Median	Technical layer	90.08	0.76	0.25	68.29	63.78	59.09
	Quantile	Biological layer	49.85	9.59	0.73	64.44	60.80	55.07
	Quantile	Technical layer	90.08	0.76	0.25	68.29	63.78	59.09
KNN	Median	Biological layer	14.14	0.47	0.00	56.01	52.41	47.06
	Median	Technical layer	0.00	0.00	0.00	46.05	41.89	35.45
	Quantile	Biological layer	13.39	0.36	0.00	55.66	51.75	46.89
	Quantile	Technical layer	0.00	0.00	0.00	46.05	41.89	35.45
RF	Median	Biological layer	75.30	30.85	3.73	69.62	65.87	60.77
	Median	Technical layer	98.73	7.63	0.51	71.05	66.89	62.03
	Quantile	Biological layer	74.15	29.07	3.58	69.41	65.52	60.38
	Quantile	Technical layer	98.73	7.12	0.51	71.12	67.03	62.03

The columns of proportion of reproducible metabolites per sample pair \(>70/80/90\%\) indicate that proportion of sample pairs of replicate experiments with reproducible metabolites per sample pair greater than \(>70/80/90\%\). The columns of proportion of reproducible sample pairs per metabolite \(>70/80/90\%\) indicate that proportion of metabolites with reproducible sample pairs per metabolite greater than \(>70/80/90\%\)

ISSN: 1471-2105