中国科技网4月8日报道（张微 编译）德克萨斯大学的两个研究人员利用一种新的研究方法，基于单核苷酸多态性(SNP)的数据分析，来推断历史人口规模的波动。在他们发表在《自然遗传学》期刊的论文中，研究人员阐述了他们如何应用他们称为SNP数据阶梯图技术来揭示数万年前的人口波动。

SNP是发生于特定人群组的DNA序列的变异。科学家们利用从中读取的信息来追踪遗传标记以及随时间变化的人口波动。在这个新的尝试中，研究人员针对从包含在千人基因组计划的9个种群中获取的DNA数据，使用了一种不同的数学分析技术。没有利用预定义人口模型，两个研究人员提出了一个弹性模型，能够利用SNP频谱数据来估计不同历史时期的人口规模。

利用他们的技术，研究人员能够发现20万年前的人口瓶颈(人口规模突然大幅减少)。他们发现，大约20-3万年前欧洲的一个人口瓶颈(其中没有包含居住在芬兰的人口)。他们还发现了一个人口瓶颈是在100-20万年前的非洲。

两位研究者称他们的模型是类似于轮廓图的多期模型，这个模型可以用来计算给定频谱的可能性，能够将每个DNA序列作为一个独立位点，研究人员指出这个模型能减少计算量，并使得研究数百个个体数据成为可能，当然也能够使历史人口的估算数据更加准确。

对9个种群的分别研究中，他们创建了200个频谱，这就能绘制一个给定历史时期基于人口规模中位数的阶梯图。他们的技术揭示出在50-7万年前，所有的种群都经历了一次严重的人口规模下降，芬兰是一个特例，它在10到2万年前有一次单独的人口规模下降，这与欧洲其他国家不同，他们的人口规模下降大约发生在一万年后。研究人员认为这是由于早期芬兰将他们与欧洲其他国家分割开的原因。研究人员还指出，非洲人口在50-7万年前经历了一次急剧下降，而且在20-3万年前也经历了一次人口规模下降。

Researchers take new approach to determine historical population fluctuations

A pair of researches with the University of Texas has used a new method to infer historical human population size fluctuations based on Single Nucleotide Polymorphism (SNP) data. In their paper published in the journal Nature Genetics, the team describes how they applied a technique they call a stairway plot to SNP data to uncover human population fluctuations going back hundreds of thousands of years.

SNP is a DNA sequence variation that occurs commonly among a given population. Scientists have been using information read from it to trace lineage and to follow population fluctuations over time. In this new effort, the researchers used a different math technique on DNA data obtained from nine populations included in the 1000 Genome Project. Rather than using techniques that rely on predefined population models, the pair of researchers instead came up with a model-flexible method that allows for using SNP spectra data to estimate population size during different periods in history.

Using their technique, the researchers were able to show population bottlenecks (sudden sharp reductions in population size) going back 200,000 years. They found, for example, that there was a population bottleneck in Europe approximately 20 to 30,000 years ago (though it did not include people living in Finland). They also found a bottleneck in Africa occurred 100 to 200,000 years ago in Africa.

The two researchers describe their method as a multi-epoch model that is similar to a skyline plot in that it can be used to calculate the likelihood of a given frequency spectrum—and it allows for treating each DNA sequence as an independent locus, which they noted reduces computational load and makes it possible to study hundreds of individuals, which of course makes historical population estimating much more accurate.

For each of the nine populations studied, the duo created 200 frequency spectra which led to a stairway plot based on median population size for a given period in history. They report that their technique revealed that all of the populations that were not Africa based experienced severe drops in population between 50 and 70,000 years ago. The population in Finland stood as a different however, in that they had a separate drop between 10 and 20,000 years ago, as opposed to the rest of Europe which saw its major drop approximately ten thousand years later. The researchers suggest this was due to early Finnish people separating themselves from the rest of Europe. They note also a sharp African population drop approximately 50 to 70,000 years ago, and then again 20 to 30,000 years ago.

来源：中国科技网