Hualian: Gene analysis of complex diseases with the concept of Pathway Cluster

In the post-genome era, the emergence of microarrays allowed researchers to explore molecular mechanisms in a macroscopic view. After many efforts and resources have been devoted to finding new disease genes, many single-gene diseases have successfully identified disease-causing genes. However, in the study of complex diseases (such as high blood pressure, diabetes and some common cancers), the harvest is not as rich as expected. In most studies of complex diseases, pathogenic genes distributed on different chromosomes can be found, but they have only small to medium linkages or associations with diseases, and only a very small number of disease-causing genes. In the large population data, the link or association of the disease is still significant. Most of the disease-causing genes currently found in complex disease research are not reproducible in cross-research reports.

Complex diseases with heterogeneity and multi- sources. For example, in Dr. Perusse1 in 2004, 113 candidate genes related to human obesity were only 18 in 50 whole-genome scan studies. The genes presented consistently positively relevant reports in more than five studies. In addition, Dr. Agarwal2's comments in 2005 mentioned (as shown in Figure 1) that 25 of the 25 hypertension genes in different linkage or association studies, 9 genes were more negatively correlated in the connectivity study than positive. Related reports. Most of the 25 genes were positively correlated and negatively correlated in association studies.
In the literature, the pathogenic genes of complex diseases lack repetitiveness across studies, and several explanations are summarized. One of the most widely accepted observations is the heterogeneous nature of these multifactorial diseases. In addition, due to different definitions of various phenotypes (such as blood pressure, blood sugar) and measurement inaccuracies, environmental hazards or warranty factors (such as smoking, intake of pollutants) Factors such as the degree of exposure and differences in genetic background between different populations can mask, enhance or alter the role of genes and cause varying degrees of disease penetrance.
In short, due to the multiplicity of causes of patients with complex diseases, the effect of any one gene mutation is diluted. So, when we bring together many patients and try to compare how their genes differ from normal people, they may find different pathogenic genes, and even find genes that are not related to the disease but are related to other characteristics of the patient.

The Pathway Cluster concept currently uses similar phenotypes to reduce heterogeneity between samples in the study of complex diseases. However, the homogenization of the phenotype is not equal to the homogenization of the genotype. Furthermore, a disease may be a combination of multiple phenotypes but different origins (genes). Although this concept has been proposed in the literature, the simplified phenotypic method used by scientists is not ideal. For example, in the field of mental illness, many scholars have proposed the concept of "endophenotype", which is the concept of "inner biological phenotype". However, the method of operation proposed by them is simply to simplify (or reduce) the phenotype, for example, to reduce the anatomy, imaging, or symptom definition, and not to reduce the biochemical pathway involved in the development of the disease. "on.
The main bottleneck of this problem is that scientists do not know enough about the mechanism of disease development. Therefore, Prof. Pan Wenhan of the Academia Sinica proposed the following suggestions: In the gene expression data generated in large quantities today, the data mining method is used for cluster analysis; the data is divided into several groups. Related, but multiple groups that are not related between groups, each group may represent the performance status of one or two minority source genes, and some of his downstream genes. The resulting cohort has a high isomorphism and is close to the potential gene of the pathogen, so it can be regarded as a pointer to the "biological path bundle".
We first use the genetic epidemiology methodology to test whether these groups are hereditary, and then use the scores obtained from this group (quantitative traits) or the further cut out of 0/1 properties for gene mapping. The rate will increase greatly.

Hypertension Research Case Dr. Lin Kexuan, a special researcher of the company, used a special artificial neural network model during the laboratory service of Prof. Pan Wenhan of the Chinese Academy of Sciences, using Huan's human expression spectrum chip (HOA). , Human OneArray®) A large amount of genetic data generated to realize the concept of "biological path bundle". As shown in Figure 2, the gene data enters the model from the input to the left of the neural network. Connect to the middle hidden node by different online connections. These intermediate hidden points represent different "biological path bundles". The genes contained in each biological path bundle are determined by different weights on the line. Finally, the results of the weighting of each gene by each biological path bundle determine whether or not to send a signal that affects the output node. Finally, the result of the weighting of the signal sent by the output to each biological path bundle determines whether or not to induce hypertension. Dr. Lin detailed the methods for determining the weight of each connection and the number of biological path bundles in the literature. In view of the limited space, it will not be detailed in this article. Figure 3 shows the relationship between the biological path bundle constructed by the neural network and blood pressure. The left side of the figure shows data for hypertensive patients, and the right side shows data for normal blood pressure. From top to bottom, the top down is the systolic pressure, diastolic pressure, model output signal, model hidden point signal and the actual gene expression in the three biological path bundles. From the model hidden point signal (Fig. 3(g) and (h)), it can be seen that the three biological path bundles show different plates in hypertensive patients and normal blood pressure. The red template indicates that the biological path bundle is in an expressed condition, and the blue template indicates that the biological path bundle is in a non- (or low) performance. As can be seen from the figure, the endophenotype 1 is strongly protective, the biological path bundle 2 is weakly protective, and the biological pathway bundle 3 is strongly risky. . These three biological path bundles divide the hypertensive patients and the normal blood pressure into several groups. Among the different groups, the biological path bundles have the highest blood pressure, and the biological path bundles have a slight drop in blood pressure, including the biological path bundle. Blood pressure drops the most. That is, the three biological path bundles constructed have different roles in blood pressure regulation. Biopath clusters can also be appropriately grouped for hypertensive patients.


in conclusion
This year, Hualian Express introduced the application of gene chips in various fields. At the end of the year, we hope to use this brief introduction to bring you to the concept of bio-path clusters and expand the possibility of constructing phenotypic-related analysis modules for gene chip data. direction. Facing the heterogeneity and multi-source nature of complex diseases or phenotypes, we anticipate that the concept of this biological pathway bundle should help to simplify the face of complex diseases or phenotypes, effectively locate disease or phenotype, and assist in identifying disease-causing Genes and other factors, in order to find effective treatment guidelines or track biomarkers at an early date.

Figure 1. Statistical report of 25 hypertension genes in different associations or association studies in Dr. Agarwal's 2005 review

Figure 2. Using a special artificial neural network model with a large amount of genetic data to simulate the relationship between genes, biological pathways and hypertension.

Figure 3. Relationship between constructed biopath clusters and blood pressure and different patient groups

references:
1. Perusse L, Rankinen T, Zuberi A, Chagnon YC, Weisnagel SJ, Argyropoulos G, Walts B, Snyder EE, Bouchard C. 2005. The human obesity gene map: the 2004 update. Obes Res 13:381–490.
2. Agarwal A, Williams GH, Fisher ND. 2005. Genetics of human hypertension. Trends Endocrinol Metab 16:127–133.
3. Pan WH, Lynn KS, Chen CH, Wu YL, Lin CY, Chang HY. Using endophenotypes for pathway cluster to map complex disease genes. Genet. Epidemiol. 2006;30:143-154.
4. Lynn KS, Li LL, Lin YJ, Wang CH, Sheng SH, Lin JH, Liao W, Hsu WL, Pan WH. A neural network model for constructing endophenotypes of common complex diseases: an application to male young-onset hypertension microarray Data. Bioinformatics. 2009 Apr 15;25(8):981-8.

High Precision Distance Sensor

Laser range sensors combine the advantages of a visible red sensing beam with the increased range of a laser. The optical distance sensor devices are Class 2 lasers, support Class 1 customized. The high precision autonics distance sensor is successfully be used to connect to Arduino, Raspberry Pi, with +-1mm to 3mm high accuray, and long-range solutions (max up to 150m) for your application.

Precision Distance Sensor,Precision Laser Distance Sensor ,Precision Distance Sensor Laser,Accurate Laser Distance Sensor

Chengdu JRT Meter Technology Co., Ltd , https://www.cdtoflidarsensor.com