Search

Additional file 3: Table S3. The list of specific kinase targets screened in the human RTK phosphorylation array.

Additional file 1: Table S1. Sequences of Primers for Real-time Polymerase Chain Reaction.

Additional file 2: Table S2. Sequences of siRNAs.

In this study, the feasibility of gas chromatography (GC) and gas chromatography–ion mobility spectrometry (GC–IMS) systems for detecting peanut oil (PO) adulterated with rapeseed oil (RO) in different ratios was evaluated. In the GC analysis, the oils were categorized on the basis of fatty acid compositions. In the GC–IMS analysis, the oils were distinguished on the basis of volatile fractions. To demonstrate the high quality of the GC–IMS results, principal component analysis (PCA) and cluster analysis (CA) were conducted. The results showed that fatty acid profiles in PO exhibited higher amounts of palmitic acid (10.70 ± 0.18%), linoleic acid (32.87 ± 0.52%), and arachidic acid (2.30 ± 0.05%). RO had higher levels of linolenic acid (7.49 ± 0.36%) and erucic acid (8.02 ± 0.39%), which were not detected in PO. The linear regression analysis showed that the change in C18:2, C18:3, C20:1, and C22:1 content was highly correlated with the adulteration proportion of RO. However, only significant differences were visible in C18:3 and C20:1 when the adulteration proportion was a minimum of 5%. Satisfactory results of adulteration determination were obtained using the GC–IMS combined with the PCA and CA. Each oil type had unique characteristic compounds, and signals of some volatile compounds changed with the increase in the adulteration proportion especially at run times ranging from 100 to 300 s. For example, the intensity of peak signals of 89–92 increased with the addition of RO while peak intensity of 104 to 107 weakened. These individual signals revealed the volatile compounds responsible for the differences between PO and RO and could help to screen PO adulterated with RO. The established PCA scatter plots and CA could correctly differentiate authentic PO from adulterated samples with at least 1% of RO. Consequently, the GC–IMS shows higher efficiency and feasibility compared with the GC and may serve as a basis for the detection of adulteration of various substances.

Additional file 9: Table S4. List of signal densities of human RTK phosphorylation Antibody Array.

Additional file 9. Table S6. AI alleles in PRAD are under significant germline selection. All exonic alleles are classified into alleles with higher frequencies in cancer population and alleles with lower frequencies in cancer population. The frequencies of each class in SNPs that undergo different somatic selection are compared to the frequencies in all exonic SNPs. The significance is based on hypergeometric test P values.

Additional file 2: Table S2. Demographic and clinical characteristics and levels of CD151 mRNA expression in NSCLC tissue.

Additional file 7. Table S4. AI alleles in COAD are under significant germline selection. All exonic alleles are classified into alleles with higher frequencies in cancer population and alleles with lower frequencies in cancer population. The frequencies of each class in SNPs that undergo different somatic selection are compared to the frequencies in all exonic SNPs. The significance is based on hypergeometric test P values.

Additional file 3. Fig. S1. The inclusion and exclusion filtering of exonic SNPs for alleleic imbalance analysis in four cancer types.

Additional file 8. Table S5. AI alleles in LUAD are under significant germline selection. All exonic alleles are classified into alleles with higher frequencies in cancer population and alleles with lower frequencies in cancer population. The frequencies of each class in SNPs that undergo different somatic selection are compared to the frequencies in all exonic SNPs. The significance is based on hypergeometric test P values.

Additional file 6. Table S3. AI alleles in BRCA are under significant germline selection. All exonic alleles are classified into alleles with higher frequencies in cancer population and alleles with lower frequencies in cancer population. The frequencies of each class in SNPs that undergo different somatic selection are compared to the frequencies in all exonic SNPs. The significance is based on hypergeometric test P values.

Additional file 5. Fig. S2. The distribution of allelele frequencies (a) and the ratios of allele frequencies (b) of exonic SNPs in four cancer types.

Additional file 1. Table S1. The upper and lower boundaries for each parameter and the prior distributions used in the method.

Additional file 1: Table S1. Demographic and clinical characteristics and levels of CD151 protein expression in NSCLC tissue.

Additional file 6: Table S3. The results of Proteome Profiler Array-Human Soluble Receptor Array Non-hematopoietic panel and Common Analytes panel. (Part N and Part C).