Accurate Tree-based Missing Data Imputation and Data Fusion within the Statistical Learning Paradigm

Framework of this paper is statistical data editing, specifically how to edit or impute missing or contradictory data and how to merge two independent data sets presenting some lack of information. Assuming a missing at random mechanism, this paper provides an accurate tree-based methodology for both missing data imputation and data fusion that is justified within the Statistical Learning Theory of Vapnik. It considers both an incremental variable imputation method to improve computational efficiency as well as boosted trees to gain in prediction accuracy with respect to other methods. As a result, the best approximation of the structural risk (also known as irreducible error) is reached, thus reducing at minimum the generalization (or prediction) error of imputation. Moreover, it is distribution free, it holds independently of the underlying probability law generating missing data values. Performance analysis is discussed considering simulation case studies and real world applications.     

Mixtures of Autoregressions with an Improper Component for Panel Data

An EM algorithm for fitting mixtures of autoregressions of low order is constructed and the properties of the estimators are explored on simulated and real datasets. The mixture model incorporates a component with an improper density, which is intended for outliers. The model is proposed as an alternative to the search for the order of a single-component autoregression. The methods can be adapted to other patterns of dependence in panel data. An application to the monthly records of income of the outlets of a retail company is presented.     

Mutations in the human homologue of mouse dl cause autosomal recessive and dominant hypohidrotic ectodermal dysplasia.

X-linked hypohidrotic ectodermal dysplasia results in abnormal morphogenesis of teeth, hair and eccrine sweat glands. The gene (ED1) responsible for the disorder has been identified, as well as the analogous X-linked gene (Ta) in the mouse. Autosomal recessive disorders, phenotypically indistinguishable from the X-linked forms, exist in humans and at two separate loci (crinkled, cr, and downless, dl) in mice. Dominant disorders, possibly allelic to the recessive loci, are seen in both species (ED3, Dlslk). A candidate gene has recently been identified at the dl locus that is mutated in both dl and Dlslk mutant alleles. We isolated and characterized its human DL homologue, and identified mutations in three families displaying recessive inheritance and two with dominant inheritance. The disorder does not map to the candidate gene locus in all autosomal recessive families, implying the existence of at least one additional human locus. The putative protein is predicted to have a single transmembrane domain, and shows similarity to two separate domains of the tumour necrosis factor receptor (TNFR) family.     

Mutations in a new gene encoding a thiamine transporter cause thiamine-responsive megaloblastic anaemia syndrome.

Thiamine-responsive megaloblastic anaemia syndrome (TRMA; MIM 249270) is an autosomal recessive disorder with features that include megaloblastic anaemia, mild thrombocytopenia and leucopenia, sensorineural deafness and diabetes mellitus. Treatment with pharmacologic doses of thiamine ameliorates the megaloblastic anaemia and diabetes mellitus. A defect in the plasma membrane transport of thiamine has been demonstrated in erythrocytes and cultured skin fibroblasts from TRMA patients. The gene causing TRMA was assigned to 1q23.2-q23.3 by linkage analysis. Here we report the cloning of a new gene, SLC19A2, identified from high-through-put genomic sequences due to homology with SLC19A1, encoding reduced folate carrier 1 (refs 8-10). We cloned the entire coding region by screening a human fetal brain cDNA library. SLC19A2 encodes a protein (of 497 aa) predicted to have 12 transmembrane domains. We identified 2 frameshift mutations in exon 2. a 1-bp insertion and a 2-bp deletion, among four Iranian families with TRMA. The sequence homology and predicted structure of SLC19A2, as well as its role in TRMA, suggest that its gene product is a thiamine carrier, the first to be identified in complex eukaryotes.     

A single EFEMP1 mutation associated with both Malattia Leventinese and Doyne honeycomb retinal dystrophy.

Malattia Leventinese (ML) and Doyne honeycomb retinal dystrophy (DHRD) refer to two autosomal dominant diseases characterized by yellow-white deposits known as drusen that accumulate beneath the retinal pigment epithelium (RPE). Both loci were mapped to chromosome 2p16-21 (refs 5,6) and this genetic interval has been subsequently narrowed. The importance of these diseases is due in large part to their close phenotypic similarity to age-related macular degeneration (AMD), a disorder with a strong genetic component that accounts for approximately 50% of registered blindness in the Western world. Just as in ML and DHRD, the early hallmark of AMD is the presence of drusen. Here we use a combination of positional and candidate gene methods to identify a single non-conservative mutation (Arg345Trp) in the gene EFEMP1 (for EGF-containing fibrillin-like extracellular matrix protein 1) in all families studied. This change was not present in 477 control individuals or in 494 patients with age-related macular degeneration. Identification of this mutation may aid in the development of an animal model for drusen, as well as in the identification of other genes involved in human macular degeneration.