1. 程式人生 > >Building a Global Network for Genomic Data – DNAnexus, an Advanced APN Technology Partner

Building a Global Network for Genomic Data – DNAnexus, an Advanced APN Technology Partner

Today’s announcement of the precisionFDA platform is significant for the genomics research community for a number of reasons. With this pilot platform, a component of President Obama’s Precision Medicine Initiative, the FDA is working towards establishing a community of stakeholders to help drive the standard around secondary analysis, the area of mapping, alignment, and variant calling in genomics research. Secondary analysis allows researchers to see variations from a given individual’s genomic makeup as compared to a reference genome. DNAnexus, an Advanced APN Technology Partner, was selected by the FDA to both power the precisionFDA platform and to build a genomics community to gather and publish reference data analysis pipelines, and reference datasets for the validation of genomic tests. The DNAnexus Platform, which was built on AWS, will deliver precisionFDA and provide the underlying cloud-based compute resources and data management.

Today I’d like to tell you about DNAnexus and its mission, what the company has built on AWS, how the cloud is helping change the face of genomics research, and DNAnexus’ work with the FDA.

Who is DNAnexus?

DNAnexus, based in Mountain View, California, has created the global network for genomics by providing an API-based platform for the sharing and management of data and tools that accelerate genomic research. The company’s mission is to deliver to organizations a secure and trusted genome informatics and data management platform, to enable organizations to tackle some of the most complex bioinformatic challenges, and to power a global network for sharing genomic information and advancing research and healthcare. The DNAnexus Platform enables scientists and clinicians worldwide to accelerate medical advances, improve patient care, and advance R&D in areas such as cancer, heart disease, Alzheimer’s disease, prenatal testing, and agricultural production. The company is an

AWS Life Sciences Competency Partner.

Why AWS?

There were a number of reasons that DNAnexus chose to build its platform on AWS, including the broad and deep security and compliance profile of AWS. “At the heart of everything we do is security and compliance,” explains Omar Serang, Chief Cloud Officer at DNAnexus. DNAnexus complies with ISO 27001 and 27002 international security standards, which ensures the highest levels of compliance with clinical regulations. “Our platform is fully capable of operating in a clinical environment under CLIA, and we can support the management of PHI data under HIPAA regulations as well,” says Serang. “We find that our security and compliance posture is a huge advantage for us in the market.”

DNAnexus leverages a number of AWS Services, and takes advantage of the global presence of AWS. “We leverage Amazon S3 and Amazon EC2 heavily, and we tend to think of S3 as the center of our universe,” explains Serang. “When you have an object store like S3 that can deliver massive payloads to hundreds and hundreds of EC2 instances without blinking, it’s an amazing capability for us. Essentially, it takes storage performance management completely off the table for us and for our customers.” The company is currently building out in the Beijing region, and has a significant presence in the EMEA, APAC, and North American regions.

The Significance of AWS for Genomics Research

“The cloud is uniquely positioned for genomic research because of the very large scales of data involved with Next Generation Sequencing (NGS) data; the massive dataset sizes from NGS are a very good fit for the cloud,” says Serang. “Additionally, the collaborative nature of genomic science is enabled and enhanced by the accessibility of the cloud.”

According to Serang, the cloud, and particularly AWS, has changed the approach researchers are able to take in genomic research. “The cloud fuels collaboration and enables people to do science that wasn’t possible before,” explains Serang. “Prior to the cloud, researchers sent hard drives around in an attempt to collaborate. We now have a different paradigm where we put data in the cloud and bring the scientists to the data, side-by-side with the EC2 compute resources they need to perform their analysis. AWS is enabling new science, and reducing the turnaround time on many different types of analysis.”

precisionFDA

Earlier this year, DNAnexus began discussions around the capabilities of its platform with Dr. Taha Kass-Hout of the FDA. Within six months, the company went from an initial concept of how the DNAnexus Platform could support precisionFDA to a contract. “It’s been a real pleasure working with the FDA scientists, particularly Dr. Kaas-Hout,” says Serang. “I feel Dr. Kaas-Hout is an absolute visionary leader. His vision of community involvement in genomics is incredibly exciting to us.” The DNAnexus Platform has an API, and on top of the API DNAnexus is delivering a web portal that is to be used by the precisionFDA community. “The web portal encapsulates fit-for-function features to be used. All of the genomic analysis, data processing, collaboration and sharing is taking place using the features of the DNAnexus platform,” explains Serang.

The first stage of the precisionFDA project is focused on community engagement. “The overall objective of the pilot is to establish a community of stakeholders around the standardization of secondary analysis, and to get this community to participate in the creation of standards,” says Serang. “The first stage is about stimulating the community to step forward and help drive these standards.” As the project develops, Serang feels that the precisionFDA platform will level the playing field and create opportunities for smaller diagnostic test providers to get tests validated and vetted. “It’s expected that a much broader community of members will be able to benefit from this platform,” explains Serang. “By having access to cloud compute and storage through an extension of the proven DNAnexus Platform, it’s extending the analytic pipelines for smaller diagnostic companies who often do not have the bioinformatic capabilities in-house and can now access best practices tools and references datasets on the precisionFDA platform. It will allow a wider range of companies to participate in the validation process, and ultimately the certification process for NGS-based diagnostics.”

To learn more about DNAnexus, visit the company’s website. Learn more about the precisionFDA project here.

You can read more about DNAnexus’s involvement in precisionFDA on their blog here.

相關推薦

Building a Global Network for Genomic DataDNAnexus, an Advanced APN Technology Partner

Today’s announcement of the precisionFDA platform is significant for the genomics research community for a number of reasons. With this pilot plat

Bioconductor(Bioconductor for Genomic Data Science教程)

mic arc nbsp nba for hub 教程 enc 文件 Bioconductor for Genomic Data Science ftp://ftp.ncbi.nlm.nih.gov/genomes/archive/old_genbank/Bacteri

A passionate advocate for open data

Radha Mastandrea wants to know what the universe is made of. More specifically, she wants to know about tiny pieces of it called quarks, the particles tha

Building a Text Editor for a Digital

The interesting exception to this tree-like structure lies in the way paragraphnodes codify their text. Consider a paragraph consisting of the sentence, “T

Show HN: Gymmmr, a social network for finding a workout partner

http://www.gymmmr.com/Gymmmr is a social network that enables people to find partners and friends to workout with. Users enter their diet information, goal

Building a CI system for Go, with Jenkins

Before continuing, why Checkout Stage ?Well, just because we’re triggering the build whenever a change is pushed to BitBucket, Jenkins is smart enough to c

caffe 教程 Fine-tuning a Pretrained Network for Style Recognition下載資料

問題:執行python examples/finetune_flickr_style/assemble_data.py --workers=1 --images=2000 --seed 831486命令下載Flickr Style資料,然而提示:Writing

ECCV2018 | 論文閱讀DetNet: A Backbone network for Object Detection

持續更新~~~ 目前大部分的目標檢測網路,包括one-stage和two-stage法,都是直接對用於影象分類的ImageNet預訓練模型進行微調,很少有專門為目標檢測設計的特徵提取器。更重要的是,影象分類和物體檢測之間存在許多差異: (i)最新的物體探測器如FPN和RetinaNet通

Pipeline Frameworks for Genomic Data

Similarly, genomic data can be passed through special software pipelines to refine and analyze the data as required, while resulting in desired visualizati

Do more with Data: Building a Data Supplier plugin for Sketch

In Sketch 52, we introduced an exciting new feature —Data. If you still haven’t read about it, be sure to check our release blog post, or take a look at th

論文筆記-DeepFM: A Factorization-Machine based Neural Network for CTR Prediction

contain feature 比較 san date res 離散 edi post 針對交叉(高階)特征學習提出的DeepFM是一個end-to-end模型,不需要像wide&deep那樣在wide端人工構造特征。 網絡結構: sparse feature

《Kalchbrenner N, Grefenstette E, Blunsom P. A convolutional neural network for modelling sentences》

概率分布 通過 AD 最小 當前 最大化 gradient function thml Kalchbrenner’s Paper Kal的這篇文章引用次數較高,他提出了一種名為DCNN(Dynamic Convolutional Neural Network)的網絡模型,在

【論文翻譯】中英對照翻譯--(Attentive Generative Adversarial Network for Raindrop Removal from A Single Image)

【開始時間】2018.10.08 【完成時間】2018.10.09 【論文翻譯】Attentive GAN論文中英對照翻譯--(Attentive Generative Adversarial Network for Raindrop Removal from A Single Imag

Six golden A Global Leader in Industrial IoT rules for creating the ideal German cover letter and r

www.inhandnetworks.de Applying for jobs is never simple but it can feel even more difficult in a foreign country when you’re unfamiliar with the l

SSR-Net: A Compact Soft Stagewise Regression Network for Age Estimation

逐級迴歸的年齡估計 本文是國立臺灣大學發表的一篇依據人臉圖片進行年齡估計的文章.受DEX論文的啟發,這篇文章也把迴歸問題轉換為多個分類問題.SSR-Net採用了由粗到細多級分類的方式.每個stage僅對其之前的預測做出更精細的判斷.因此,對於神經元個數的需要就大大減少,這樣模型的體積也就下來了.

人臉對齊(二十一)--A Recurrent Encoder-Decoder Network for Sequential Face Alignment

轉自:https://blog.csdn.net/shuzfan/article/details/52438910 本次介紹一篇關於人臉關鍵點檢測(人臉對齊)的文章: 《ECCV16 A Recurrent Encoder-Decoder Network for Sequential Fac

MSCNN論文解讀-A Unified Multi-scale Deep Convolutional Neural Network for Fast Object Detection

 多尺度深度卷積神經網路進行快速目標檢測: 兩階段目標檢測器,與faster-rcnn相似,分為an object proposal network and an accurate detection network. 文章主要解決的是目標大小不一致的問題,尤其是對小目標的檢測,通過多

LiveScan3D: A Fast and Inexpensive 3D Data Acquisition System for Multiple Kinect v2 Sensors

LiveScan3D:用於多個Kinect v2感測器的快速、低成本的3D資料採集系統 文章翻譯 引言:我們提出了一種利用多個Kinect v2感測器進行實時3D採集的方法。與使用單個感測器的方法不同,比如[1],我們可以同時記錄多個視點的動態場景。     我

A Convolutional Neural Network for Modelling Sentences

引言 Nal Kalchbrenner等人在2014年arXiv上的paper,原文地址:arXiv:1404.2188v1 [cs.CL] 8 Apr 2014。 自然語言處理的基礎問題在於句子的語義表示,其他特定的任務如分類等都是在語義表示的基礎上進行高層次的處理,所以如何對句子

Attentive Generative Adversarial Network for Raindrop Removal from A Single Image論文理解

  概述: 在去雨的過程中給網路加上了attention提取,讓網路能夠更好地學到有雨滴部分的差別。 網路結構如下: 首先使用attention提取網路來獲得包含雨滴的影象的attention影象(值在0-1之間,包含雨滴的地方值較大),attention提取網路中使用通