IUPAC project: a glossary of concepts and terms in chemometrics.

A project has been initiated by the International Union of Pure and Applied Chemistry (IUPAC) to create a glossary of concepts and terms in chemometrics. This will be accomplished by consultation with the community through the means of a wiki--a web site that can be modified by users (see http://www.iupacterms.eigenvector.com/index.php?title=Main_Page). Over time new terms can be added, and consensus definitions arrived at. The definitions will be published as IUPAC recommendations.

How to avoid over-fitting in multivariate calibration--the conventional validation approach and an alternative.

This paper critically reviews the problem of over-fitting in multivariate calibration and the conventional validation-based approach to avoid it. It proposes a randomization test that enables one to assess the statistical significance of each component that enters the model. This alternative is compared with cross-validation and independent test set validation for the calibration of a near-infrared spectral data set using partial least squares (PLS) regression. The results indicate that the alternative approach is more objective, since, unlike the validation-based approach, it does not require the use of 'soft' decision rules. The alternative approach therefore appears to be a useful addition to the chemometrician's toolbox.

Selecting the optimum number of partial least squares components for the calibration of attenuated total reflectance-mid-infrared spectra of undesigned kerosene samples.

Selecting the correct dimensionality is critical for obtaining partial least squares (PLS) regression models with good predictive ability. Although calibration and validation sets are best established using experimental designs, industrial laboratories cannot afford such an approach. Typically, samples are collected in an (formally) undesigned way, spread over time and their measurements are included in routine measurement processes. This makes it hard to evaluate PLS model dimensionality. In this paper, classical criteria (leave-one-out cross-validation and adjusted Wold's criterion) are compared to recently proposed alternatives (smoothed PLS-PoLiSh and a randomization test) to seek out the optimum dimensionality of PLS models. Kerosene (jet fuel) samples were measured by attenuated total reflectance-mid-IR spectrometry and their spectra where used to predict eight important properties determined using reference methods that are time-consuming and prone to analytical errors. The alternative methods were shown to give reliable dimensionality predictions when compared to external validation. By contrast, the simpler methods seemed to be largely affected by the largest changes in the modeling capabilities of the first components.

Prediction of pork quality attributes from near infrared reflectance spectra.

Near infrared spectroscopy (NIRS) is one of the most promising techniques for large-scale meat quality evaluation. We investigated the potential of NIRS-based models to predict drip loss and shear force of pork samples. Near infrared reflectance spectra (1000-2500 nm), water-holding capacity, shear force, ultimate pH, and colour (L(∗), a(∗), b(∗)-value) of 96 pork longissimus muscles were recorded at 2 days post mortem. Stepwise multiple linear regression (SMLR) and partial least squares regression (PLSR) analyses were used to formulate models for drip loss and shear force. Prediction models for drip loss correlated moderately strong with measured drip loss (R=0.71-0.74), which is similar to the correlation obtained using a combination of ultimate pH, filter paper test, and L(∗)-value (R=0.74). The current results indicate that NIRS enables the classification of pork longissimus muscles with a superior or inferior water-holding capacity as having a drip loss lower than 5% or higher than 7%. No useful models could be constructed for shear force.

Towards a rehabilitation of the generalized rank annihilation method (GRAM).

The trilinear PARAFAC model occupies a central place in multiway analysis, because the components of a data array can often be uniquely resolved. This paper compares the resolution for a large variety of methods, namely the generalized rank annihilation method (GRAM), alternating least squares (ALS), alternating trilinear decomposition (ATLD), alternating coupled vectors resolution (ACOVER), alternating slice-wise diagonalization (ASD), alternating coupled matrices resolution (ACOMAR), self-weighted alternating trilinear decomposition (SWATLD), and pseudo alternating least squares (PALS). The comparison was conducted using Monte Carlo simulations. It was shown that GRAM performs well for moderately and highly overlapped data. These results argue strongly against the previously claimed superiority of the alternatives listed above.

Determination of human platelet antigen frequencies in the Dutch population by immunophenotyping and DNA (allele-specific restriction enzyme) analysis.

Platelets from 200 random Dutch blood donors were typed for the human platelet alloantigens HPA-1 to -5 recognized at present and for Naka. Naka is an epitope on glycoprotein IV, not expressed on the platelet of individuals with hereditary GP IV deficiency. Platelet immunofluorescence and monoclonal antibody-specific immobilization of platelet antigens (MAIPA) were applied for this purpose. The observed phenotype frequencies were 97.86% and 28.64% for HPA-1a and -1b, 100% and 13.15% for HPA-2a and -2b, 80.95% and 69.84% for HPA-3a and -3b, 100% and 0% for HPA-4a and -4b, 100% and 19.7% for HPA-5a and HPA-5b, respectively. Platelets from all donors reacted with the anti-Naka antibodies. To determine the gene frequencies for the HPA-1, HPA-2 and HPA-3 systems directly, DNA from 98 of these donors was isolated from peripheral blood mononuclear leucocytes and specific fragments were amplified by polymerase chain reaction (PCR). The fragments were analyzed using allele-specific restriction enzymes (ASRA). In all amplified PCR products an "internal control" for each assay, ie, a restriction site for the applied enzyme independent from the phenotype of the donor was present. In all donors tested, phenotypes, as determined by serological methods and genotypes, directly determined by the ASRA, were identical. Thus, the PCR-ASRA described in this report is a practical and reliable technique for the determination of alleles that code for platelet antigen allotypes, at least in the Dutch population.

Single point mutation in human glycoprotein IIIa is associated with a new platelet-specific alloantigen (Mo) involved in neonatal alloimmune thrombocytopenia.

Here we describe a new platelet-specific alloantigen that was identified in a case of neonatal alloimmune thrombocytopenia. This antigen has provisionally been called "Mo." By studying the Mo family, it was shown to be inherited in an autosomal dominant manner. Immunoprecipitation and Western blot analysis showed that the antigen resides on platelet glycoprotein IIIa (GP IIIa). Genomic analysis, performed by applying polymerase chain reaction and sequencing, showed a C-->G substitution of base pair 1267 of the coding region of the DNA for GP IIIa, resulting in a substitution of Proline407 by Alanine407. That this substitution is associated with the antigen could be demonstrated by restriction fragment length polymorphism analysis of cDNA, prepared from platelet RNA, and of genomic DNA. It was confirmed by dot-blot hybridization with allele-specific oligonucleotides. All family members, also those being Mo antigen-positive, were healthy. None of them appeared to suffer from increased tendency of bleeding or thrombosis. Thus, the Mo mutation does not lead to significant platelet dysfunction in vivo with heterozygous carriers. One of 450 random healthy blood donors who were tested was positive for the Mo antigen. Typing was performed by the classical serologic methods as well as by DNA analysis.

NH2-terminal globular domain of human platelet glycoprotein Ib alpha has a methionine 145/threonine145 amino acid polymorphism, which is associated with the HPA-2 (Ko) alloantigens.

The glycoprotein (GP) Ib/IX complex, a prominent platelet GP complex, is the primary receptor for vWF. Previously, we have established that an antigenic polymorphism of platelets, the HPA-2 or Ko alloantigen system, is located on the 45-kD amino-terminal globular domain of GPIb alpha. With the polymerase chain reaction, we have amplified two segments of the GPIb alpha gene coding for the first 382 amino acids of two HPA-2a and two HPA-2b homozygous individuals. Nucleotide sequence analysis revealed as the only difference a C-T polymorphism at position 434 of the coding region for the mature protein. This base change results in a substitution of threonine (ACG) in HPA-2a (Kob) to methionine (ATG) in HPA-2b (Koa) at amino acid position 145. The C-T polymorphism is reflected in a difference in restriction enzyme recognition, resulting in an Aha 2-site in the HPA-2b allele and a SfaN1 site in the HPA-2a allele. Restriction fragment length polymorphism analysis of the amplified DNA of 3 HPA-2(a-,b+), 2 HPA-2(a+,b+), and 11 HPA-2(a+,b-) donors showed that these restriction sites were associated with the HPA-2 alleles. DNA-typing for the HPA-2 alloantigen system on genomic DNA obtained from a small number of cells may be applied for determining the genotype of a fetus from an immunized mother or of severely thrombocytopenic patients.