Inference of multiple-wave population admixture by modeling decay of linkage disequilibrium with polynomial functions.

To infer the histories of population admixture, one important challenge with methods based on the admixture linkage disequilibrium (ALD) is to remove the effect of source LD (SLD), which is directly inherited from source populations. In previous methods, only the decay curve of weighted LD between pairs of sites whose genetic distance were larger than a certain starting distance was fitted by single or multiple exponential functions, for the inference of recent single- or multiple-wave admixture. However, the effect of SLD has not been well defined and no tool has been developed to estimate the effect of SLD on weighted LD decay. In this study, we defined the SLD in the formularized weighted LD statistic under the two-way admixture model and proposed a polynomial spectrum (p-spectrum) to study the weighted SLD and weighted LD. We also found that reference populations could be used to reduce the SLD in weighted LD statistics. We further developed a method, iMAAPs, to infer multiple-wave admixture by fitting ALD using a p-spectrum. We evaluated the performance of iMAAPs under various admixture models in simulated data and applied iMAAPs to the analysis of genome-wide single nucleotide polymorphism data from the Human Genome Diversity Project and the HapMap Project. We showed that iMAAPs is a considerable improvement over other current methods and further facilitates the inference of histories of complex population admixtures.

Automatic discovery of sub-molecular sequence domains in multi-aligned sequences: a dynamic programming algorithm for multiple alignment segmentation.

Automatic identification of sub-structures in multi-aligned sequences is of great importance for effective and objective structural/functional domain annotation, phylogenetic treeing and other molecular analyses. We present a segmentation algorithm that optimally partitions a given multi-alignment into a set of potentially biologically significant blocks, or segments. This algorithm applies dynamic programming and progressive optimization to the statistical profile of a multi-alignment in order to optimally demarcate relatively homogenous sub-regions. Using this algorithm, a large multi-alignment of eukaryotic 16S rRNA was analyzed. Three types of sequence patterns were identified automatically and efficiently: shared conserved domain; shared variable motif; and rare signature sequence. Results were consistent with the patterns identified through independent phylogenetic and structural approaches. This algorithm facilitates the automation of sequence-based molecular structural and evolutionary analyses through statistical modeling and high performance computation.

CLIFF: clustering of high-dimensional microarray data via iterative feature filtering using normalized cuts.

We present CLIFF, an algorithm for clustering biological samples using gene expression microarray data. This clustering problem is difficult for several reasons, in particular the sparsity of the data, the high dimensionality of the feature (gene) space, and the fact that many features are irrelevant or redundant. Our algorithm iterates between two computational processes, feature filtering and clustering. Given a reference partition that approximates the correct clustering of the samples, our feature filtering procedure ranks the features according to their intrinsic discriminability, relevance to the reference partition, and irredundancy to other relevant features, and uses this ranking to select the features to be used in the following round of clustering. Our clustering algorithm, which is based on the concept of a normalized cut, clusters the samples into a new reference partition on the basis of the selected features. On a well-studied problem involving 72 leukemia samples and 7130 genes, we demonstrate that CLIFF outperforms standard clustering approaches that do not consider the feature selection issue, and produces a result that is very close to the original expert labeling of the sample set.

Molecular alterations of p73 in human esophageal squamous cell carcinomas: loss of heterozygosity occurs frequently; loss of imprinting and elevation of p73 expression may be related to defective p53.

p73 is structurally and functionally related to p53 and is possibly a tumor suppressor gene. Using 15 surgically resected frozen esophageal specimens containing both squamous cell carcinomas (ESCC) and neighboring normal epithelia, we studied p73 gene alterations and mRNA expression. Loss of heterozygosity of the p73 loci was found in nine of 14 informative cases (64%). A polymorphism at codon 173 (Thr) of p73 was identified (eight samples had ACC and seven had ACT), but mutation was not detected in tumor samples. Nine of the 15 ESCC samples (60%) displayed significantly elevated expression of p73 over the neighboring normal epithelium; of these nine samples, four displayed loss of imprinting (LOI) and one switched the expressed allele. Hypermethylation of exon 1 of the p73 gene was not detected, using the bisulfite modification method, in normal or tumor samples. Twelve of the 15 (80%) ESCC samples contained p53 defects, including missense mutation, non-frameshift small deletion or insertion, non-detectable transcripts and protein accumulation. The ESCC samples with p53 defects were significantly correlated with those which had elevated expression of p73 (Fisher's exact test, P < 0.05). The results suggest that increased expression of p73, including that by LOI, could be a partial compensatory mechanism for defective p53.

Mechanisms of inactivation of p14ARF, p15INK4b, and p16INK4a genes in human esophageal squamous cell carcinoma.

The 9p21 gene cluster, harboring growth suppressive genes p14ARF, p15INK4b, and p16INK4a, is one of the major aberration hotspots in human cancers. It was shown that p14ARF and p16INK4a play active roles in the p53 and Rb tumor suppressive pathways, respectively, and p15INK4b is a mediator of the extracellular growth inhibition signals. To elucidate specific targets and aberrations affecting this subchromosomal region, we constructed a detailed alteration map of the 9p21 gene cluster by analyzing homozygous deletion, hypermethylation, and mutation of the p14ARF, p15INK4b, and p16INK4a genes individually in 40 esophageal squamous cell carcinomas (ESCCs) and compared the genetic alterations with mRNA expression in 18 of these samples. We detected aberrant promoter methylation of the p16INK4a gene in 16 (40%), of p14ARF in 6 (15%), and of p15INK4b in 5 (12.5%) tumor samples. Most p16INK4a methylations were exclusive, whereas all but one of the p14ARF/p15INK4b methylations were accompanied by concomitant p16INK4a methylation. We detected homozygous deletion of p16INK4a in 7 (17.5%), of p14ARF-E1beta in 13 (33%), and of p15INK4b in 16 (40%) tumor samples. Most deletions occurred exclusively on the E1beta-p15INK4b loci. Two samples contained p14ARF deletion but with p16INK4a and p15INK4b intact. No mutation was detected in the p14ARF and p16INK4a genes. Comparative RT-PCR showed good concordance between suppressed mRNA expression and genetic alteration for p15INK4b and p16INK4a genes in the 18 frozen samples, whereas 5 of the 13 cases with suppressed p14ARF mRNA expression contained no detectable E1beta alteration but aberrations in the p16INK4a locus. Our results show that in human ESCCs, p14ARF is a primary target of homozygous deletion along with p15INK4b, whereas p16INK4a is the hotspot of hypermethylation of the 9p21 gene cluster. The frequent inactivation of the p14ARF and p16INK4a genes may be an important mechanism for the dysfunction of both the Rb and p53 growth regulation pathways during ESCC development.

Loss of heterozygosity of the Rb gene correlates with pRb protein expression and associates with p53 alteration in human esophageal cancer.

To understand the alterations of Rb tumor suppressor gene and the relationship between defects in the Rb and p53 pathways in human esophageal carcinogenesis, we examined the loss of heterozygosity (LOH) of the Rb gene and immunohistochemical staining of pRb protein in 56 esophageal squamous cell carcinoma specimens and related the results to the p53 gene alterations. Using four introgenic polymorphic markers as probes, we observed LOH of the Rb gene in 30 of the 55 informative tumor samples. Immunohistochemical analysis revealed different patterns of pRb expression among the tumor samples. In the 56 cases, 16 displayed extensive pRb staining comparable to that of the adjacent normal epithelia, whereas 33 showed either significantly decreased or no pRb staining and 7 had a focal staining pattern reflecting heterogeneous cancer nests in the tumor with respect to Rb status. In the tumor samples containing Rb LOH, 90% showed low or no pRb expression, whereas in samples without Rb LOH, only 20% had altered pRb expression. There was a strong association between LOH of the Rb gene and alteration of pRb expression in our samples (P < 0.0001), suggesting LOH is a main event leading to Rb inactivation. We found that Rb LOH was more frequent in tumors with p53 mutations (P < 0.05), which occurred in 31 of the 49 cases analyzed. When the status of Rb and p53 alterations was evaluated by the combined results of immunohistochemical and genetic analyses, we found that alteration of Rb and p53 had an even stronger association in our esophageal squamous cell carcinoma samples (P = 0.0015). Among the 51 cases in which both the Rb and p53 status were determined, 31 contained alterations in both genes, and only 5 and 6 cases were altered in only Rb and only p53, respectively. Our results suggest that defects in the Rb and p53 pathways and their potential synergistic effect in deregulating cell cycle and apoptosis are major mechanisms for esophageal carcinogenesis.

Role of p53 gene mutations in human esophageal carcinogenesis: results from immunohistochemical and mutation analyses of carcinomas and nearby non-cancerous lesions.

In order to characterize p53 alterations in esophageal cancer and to study their roles in carcinogenesis, we performed gene mutation and immunohistochemical analysis on 43 surgically resected human esophageal specimens, which contain squamous cell carcinoma (SCC) and adjacent non-cancerous lesions, from a high-incidence area of Linzhou in Henan, China. A newly developed immunohisto-selective sequencing (IHSS) method was used to enrich the p53 immunostain-positive cells for mutation analysis. p53 gene mutations were detected in 30 out of 43 (70%) SCC cases. Among 29 SCC cases that were stained positive for p53 protein, 25 (86%) were found to contain p53 mutations. In five cases of SCC with homogeneous p53 staining, the same mutation was observed in samples taken from four different positions of each tumor. In a well differentiated cancer nest, p53 mutation was detected in only the peripheral p53-positive cells. In tumor areas with heterogeneous p53 staining, either the area stained positive for p53 had an additional mutation to the negatively stained area or both areas lacked any detectable p53 mutation. In the p53-positive non-cancerous lesions adjacent to cancer, p53 mutations were detected in seven out of 16 (47%) samples with basal cell hyperplasia (BCH), eight out of 12 (67%) samples with dysplasia (DYS), and six out of seven (86%) samples with carcinoma in situ (CIS). All mutations found in lesions with DYS and CIS were the same as those in the nearby SCC. In seven cases of BCH containing mutations, only three had the same mutations as the nearby SCC. The results suggest that p53 mutation is an early event in esophageal carcinogenesis occurring in most of the DYS and CIS lesions, and cells with such mutations will progress to carcinoma, whereas the role of p53 mutations in BCH is less clear.

Aberrant methylation of p16INK4a and deletion of p15INK4b are frequent events in human esophageal cancer in Linxian, China.

p16INK4a and p15INK4b genes, which encode two functionally related CDK inhibitors, recently emerged as candidate tumor suppressor genes since they were both localized to 9p21, which frequently undergoes hemizygous and homozygous deletion in a variety of tumor types. To determine the mode of inactivation of these two genes in human esophageal squamous cell carcinoma (ESCC), we performed multiple molecular analyses in 60 ESCC specimens from Linxian, China using DNA methylation assay, LOH analysis, deletion screening and SSCP-sequencing. We observed that p16INK4a inactivation was predominantly associated with aberrant methylation in the CpG island of its promoter region, whereas p15INK4b frequently had homozygous deletions. Compared with aberrant methylation, which occurred in 17 of 34 cases, homozygous deletion of p16INK4a and LOH at its nearby D9S942 microsatellite marker were observed at a much lower frequency (17%). Intragenic mutation in p16INK4a gene was rare. In contrast, homozygous deletion in p15INK4b and LOH at the nearby D9S171 marker were observed at frequencies of 35 and 47%, respectively, and the two events were significantly associated with each other. On the other hand, aberrant methylation of p15INK4b was relatively infrequent (6/34) and occurred concomitantly with p16INK4a methylation. Among the 60 cases, only four contained a continuous homozygous deletion spanning both p15INK4b and p16INK4a. Six cases were exclusively deleted at p16INK4a and 17 exclusively deleted at p15INK4b. LOH at D9S942 and D9S171 was also found to be mutually exclusive. Our results suggest that the alteration mode at 9p21 was not uniform, and the two genes were inactivated by distinct mechanisms. Altogether, 68% of the samples harbor at least one type of alteration in p16INK4a gene and 50% of the samples were altered in p15INK4b gene, indicating that they are the frequent inactivating targets during ESCC development.