Video在线视频

我是Adam Witney
-I am Adam Witney

来自英国伦敦大学圣乔治学院的
from the Institute for Infection and Immunity

感染与免疫研究院
of St George's University of London, UK.

我将谈谈通过基因组测序进行耐药性预测
I will talk about drug resistance prediction by genome sequencing.

一旦从选择的平台产生读数
Once we have generated the reads from the chosen platform,

分析的第一步
the first step of the analysis

是比对原始序列读数与参考序列
is to compare the raw sequence reads against the reference sequence.

通常
Very often,

参考序列是全基因组参考序列
the reference sequences are whole-genome references

在结核病中
and in the case of TB,

通常是分离株H37Rv的序列
this is commonly the H37Rv isolate sequence

或是与已知的对特定种类药物耐药的
or alternatively reference sequences associated with genes

基因相关的参考序列
which are known to confer resistance to specific classes of drugs can be used.

例如螺旋酶A基因与氟喹诺酮耐药性相关
For example the gyrase A gene is associated with fluoroquinolone resistance.

一旦我们将读数与参考序列进行比较和校正
Once we have compared and aligned our reads to the reference,

下一步就是
the next step

当已知位点未发生突变时
is to identify whether there are mutations

确定这些特定参考序列中是否发生突变
in those particular references and when there are no mutations at known sites.

然后将其与SNP数据库进行比较
This will then be compared against the SNP database

该数据库是与耐药性相关的一组已知突变
which is a set of known mutations associated with resistance

使我们能够进行耐药性预测
which will then enable us to make our resistance call.

两个问题影响了我们预测耐药性的可信度
Two issues affect the confidence of our ability to predict resistance.

一是序列读数的覆盖深度
The first is the depth of coverage of our sequence reads.

更多的覆盖深度使对任何特定位点的
A higher depth of coverage gives us a greater confidence of resistant mutations

耐药性突变或野生型序列可信度更高
or wild-type sequences at any particular site.

二是我们的SNP数据库
The second step is the ability of our SNP database

真正能够有把握地进行耐药性预测的能力
to actually be able to predict with confidence a resistance call.

随着分离株序列的不断增加
With ever-increasing isolate sequencing,

我们的数据库会变得更好
our databases will only get better

且增强对来自任何特定序列的
and enhance their prediction confidence of the resistance

耐药性预测的可信度
call from any particular sequence.

我提到的几个平台常用于生成读数
There are several platforms as I mentioned:

如MiSeq Ion Torrent和 MinION
MiSeq, Ion Torrent, MinION are very commonly used for generating the reads.

与参考序列进行校正的
Aligning them to references

是bwa tmap校正工具
are the bwa, tmap alignment tools.

对于调用突变
For calling mutations,

samtools和GATK调用
there are the samtools and GATK calls

可以给我们一个SNP或野生型序列的列表
which will give us a list of SNP or wild-type sequences.

结核病的常用SNP数据库
Popular SNP databases for TB

是TBDReaMDB数据库或ReseqTB数据库
are the TBDReaMDB database or the ReseqTB database.

此外
Also,

还有几个与世界各地不同组织的
there are several well-known databases associated with

研究出版物有关联的著名数据库
research publications out of various groups around the world.

在产生调用突变方面
In terms of generating the resistance calls,

有一些工具
there are tools

如TB Profiler
like TB Profiler

可完整地分析读数序列
which will run all the way from the reads

以进行耐药性预测和置信度预测
to a resistance call prediction and confidence prediction.

另外有一个名为Mykrobe Predictor的
And also a standalone tool

单机工具
called Mykrobe Predictor

它允许我们拖拽一组读数到此应用程序
which allows us to drag onto the application a set of reads

然后基于这些读数预测耐药性
which will then predict a resistance from those reads.

TB Profiler是伦敦卫生和热带医学院研发的
TB Profiler is coming out of the London School of Hygiene and Tropical medicine.

Mykrobe Predictor
Mykrobe Predictor

是EBI的Zamin Iqbal小组研发的
is coming out of Zamin Iqbal's group at the EBI.

这里有链接可以下载和试用上述软件
Here are some links to download the software tools and try them out.

感谢您的聆听
Thank you for listening.