A supervised learning approach, similar to that described in [ 4, 9], was applied to design a gene expression profile for the identification of samples with low and high TCP (see below).
Although most investigators choose to measure only total phenol and paraben concentrations (conjugated and free combined), measuring both species in at least a subgroup of study samples allows for the identification of samples in which the free:total proportion is out of the expected range (Koch et al. 2012; Völkel et al. 2002; Waechter et al. 2007), which may indicate contamination.
We applied a single cut-off at score 4, the median and mean value of the entire cohort, for the identification of samples producing low-moderate (CXCL12low/moderate, scores 0-4, n = 97) and high (CXCL12high, scores 5-7, n = 86) levels of CXCL12.
Our neighbor-joining tree allowed for the identification of 2 samples as albacore, 2 samples as bigeye tuna, 2 samples as northern bluefin tuna, and 2 samples as southern bluefin tuna (12%), but was too poorly resolved to support the identifications of any other samples (Figure 4).
The database approach consists basically of three steps: measuring reference substances to generate a database, measuring unknown or difficult samples and using the database for the identification of these samples.
Additionally, the possibility of using standard diagnostics for the identification of outlying samples proved to be particularly useful in detecting marble samples which do not correspond to any reference material and therefore prevents incorrect interpretations and assignments being made.
In order to test the suitability of the clustering optimization for the identification of Peronospora samples, we added the 14 newly sequenced downy mildew samples to the previously generated multiple sequence alignment, using the -read_msa option of the poa alignment software.
The application of DNA markers for the identification of biological samples from both human and non-human species is widespread and includes use in food authentication.
Fig. 1 depicts the practical workflow we propose for the identification of mtDNA samples that derive from closely maternally related donors.
All samples in the UK dataset were sent to a central laboratory in the ICR (Sutton/UK) and served as a positive control for the identification of shipped samples.
