Is there really a genetic algorithm?
Release date: 2013-07-12 Is there really a genetic algorithm? Genome-wide association studies have identified genetic variants associated with academic achievement, but each of these mutations has a limited impact. A genome-wide association study found mutations that might affect the complex behavior of each individual's educational attainment. Basic research on genetics of human behavior has always been a controversial research direction. For example, some people have studied both identical twins and nonidentical twins in an attempt to predict how genetic mutations can affect their differences in income and years of education (in other words When it is said that trying to study how far these people enter the university and become rich in heredity, some people think that this kind of research is meaningless, and it should prohibit the detection of heritability in social and economic achievements. And research. However, Rietveld et al. claimed that they have made some new progress in this area and discovered genes related to their academic achievements. Similar news will come out one or two at regular intervals, but in the end it is always confirmed to be false news. How should we face and interpret this scientific discovery? Like complex traits such as height and weight, behavioral traits are likely to be a trait that is affected by hundreds of thousands of small genetic components that function as a whole. Someone who has tried to find these genetic components by studying a whole family member can ultimately fail. They have searched for genetic markers of co-heritance with traits; they have also studied candidate neurobiological traits or candidate genes that may be associated with these neurobiological traits, but unfortunately none Can be successful. Genome-wide association studies can detect whether a personality is associated with one or more of the hundreds or even millions of mutations. At present, there is basically a consensus on genome-wide association studies, that is, this method is very useful for non-behavioral categories of traits, but it should be reminded that this method can only find a small number of mutation sites, and can not find All relevant sites. However, in addition to a few good luck, genome-wide association studies have almost nothing in the analysis of weight, height and certain diseases. This may be because behavior is not the same as other traits, and it is not regulated by genetics. For autism and schizophrenia, a large proportion of pathogenic genetic defects may be very rare mutations, some of which may be suddenly new, so it is impossible to pass a genome-wide association study. Discover them. Only known, relatively common mutations can be found using genome-wide association studies. In addition, the genetic effects of these mutations on behavior are very small, much smaller than we expected, and the number of such mutations is much more than we expected, so it is difficult to find these large quantities, but the role is A negligible genetic mutation. However, the research by Rietveld et al. has some gains and provided us with some answers. They found that the mutations associated with academic achievement (mainly using years of education and whether or not to complete university studies) have little effect on each mutation. The effects of these mutations are much smaller than those of physiological traits, and the effects of these mutations can only be found if the test sample is very large (a total of 126,000 people were analyzed in this study). The most influential mutations were only 0.02% effective, while the height-affecting mutations had an effect of about 0.4%, and the weight-influencing mutations had an effect of about 0.3%, compared to a whole one. Magnitude. However, this one alone is not sufficient to explain the overall effect of all these mutations. Therefore, it is particularly important to conduct another observation test. We hypothesized that only 10 genetic loci are related to academic achievement, and the signals of these 10 loci are also ranked in the top ten in genome-wide association studies, and we also assume that the influence of each locus is The largest, that is, can reach 0.02%. So if you consider all the signals of these 10 sites, think of them as a whole predictor, how much influence will this predictor have on academic achievement? In fact, this is the so-called polygenic score method. If the 10 sites we selected are all really useful sites, then their overall effect should be 10 × 0.02, which can reach 0.2%. If you enlarge this number, assume that there are hundreds of thousands. Or more sites play a role, then their effect is very large enough to explain the effect of genetics on behavioral traits. However, if only 10 sites are really effective, then there is actually no substantial impact, but it will combine the effects of other sites. As Rietveld et al. found, the greater the number of mutations involved in functioning, the more accurate the prediction of the outcome will be. This means that academic achievement seems to be really a trait affected by multiple genes. But why are the effects of these sites so small? Is it because our behavioral testing methods are too bad to detect the real impact? The work of Rietveld et al. has confirmed this to some extent. Rietveld believes that if you can make full use of various phenotypic testing methods, such as years of education, then genome-wide association studies may also play a role in helping us find genes that are involved in a broader social connection. These genes can then be further analyzed in conjunction with other relevant information, even if it is not yet certain. Genome-wide research has practical implications for discovering the methods used in candidate genes. In fact, we have found that the extent of smoking is related to mutations in the long arm of chromosome 15, which is found by investigating the number of cigarettes smoked in large numbers of people. We can perform very accurate tests on this phenotype, but the actual exposure to tobacco can only be non-quantitatively tested. This is mainly because the degree and manner of smoking varies from person to person. The person will suck the smoke deep into the lungs, and some people like to keep the smoke all the time, there is no left. Since the consumption of nicotine is most likely to be affected by genetic factors, many methods for directly detecting nicotine can more accurately detect the influence of genetic factors, and can also clarify the possible association mechanisms. The number of years of education has made Rietveld and others feel the importance of finding suitable trait detection methods. For academic achievement, we can have a lot of testing methods, sometimes a short questionnaire is completed, sometimes it will be Do a very detailed assessment. However, there is no evidence that the genetic effects detected using relatively coarse detection methods are weaker than when using precise detection methods. Although cigarette consumption is a relatively reliable measure, it does not accurately reflect actual tobacco (nicotine) exposure. Similarly, the number of years of schooling is a relatively accurate measure, but even in school, the years of education are all the same, and everyone's academic achievement is different. In Rietveld's research, there is still the question of which test indicator should be selected for research? It seems that we did find some clues, but there is more work to be done later. The vague nature of the behavioral phenotype determines that this is not an easy task, and it is far more difficult than finding genetic mutations related to smoking. According to Rietveld's research, we can think that academic achievement and years of education may partially reflect a person's intellectual level, and those with high IQs usually perform better in school. So is Rietveld's research the first case to successfully study the genetic basis of IQ? This will definitely lead to a new round of controversy. Source: lifeomics
JRT has been R&D Laser Distance Sensor every year. In this part, we put every new sensor here. To help our customer to find the lastest item. To know the updates we are doing, what's new for laser distance modules. What kind of Laser Distance Meter sensor we can make. New Laser Distance Sensor,New Laser Distance Module,New Distance Sensor,New Laser Sensor Chengdu JRT Meter Technology Co., Ltd , https://www.rangesensors.com
Best Regards
JRT Team