To ensure that our sampling approach yields sufficiently variable samples, we also quantify the overlap between samples by computing the fraction of genes shared by an increasing number of samples, using the "monogenic" set as an example.
Before preparing the taste samples, we also had the participants fill out a questionnaire about their present feelings (among other items).
To find the antireflection performance on our samples, we also conduct the reflectance measurement on SiGe nanostructures grown at different temperatures.
Through joint analysis of metabolomic data alongside clinical features of patient samples, we also identified a small number of metabolites, including several polyamines and kynurenine, which were associated with aggressive tumors across several tumor types.
Since the dielectric relaxation processes are very sensitive to prior structure of the samples, we also used differential scanning calorimetry, IR spectroscopy and gel measurements to determine the changes in crystal fraction, oxidative degradation and degree of network formation, respectively.
Because the model input data were extremely unbalanced between ignition and nonignition samples, we also introduced a cost-sensitivity analysis into the RF model (RF-cost sensitive) in addition to the standard RF model (RF-original).
For selected samples, we also conducted a comparison between the conventional single beam TOF-SIMS depth profiling and a high resolution dual beam depth profiling and found that the former approach can detect the same features in depth profiles as the latter one.
In addition to providing azo3 as a valuable tool to analyze sulfides in biological samples, we also discussed the influences of the electron effect on the sensitivity of the probes, which would shed some light on the design of future reaction-based probes.
In some samples we also found many radiolarians, haptophytes and picobiliphytes.
For reference samples, we also collected one faecal sample and several urine samples from females.
In CC erythrocyte samples, we also detected most band 3 proteins in the bottom fractions 30 36.
