Jump to content

英文维基 | 中文维基 | 日文维基 | 草榴社区

Talk:Emerald Cloud Lab

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia


source mining

[edit]
one of the first two cloud-based robotic labs (along with Transcriptic)
The first chemistry-orientic cloud lab
(only a minor description in a paragraph)
RS? Yes, this seems reliable, since this author is a real Forbes author, and Forbes is considered good for its typical articles. However, this is only a throwaway mention, so not good for notability.
Actually no. Contributors aren't on staff Jlevi (talk) 14:46, 6 November 2019 (UTC)[reply]

- "in Mountain View, California"

- "creating faster and cheaper drug experiments by putting them in the cloud"

- cofounder is DJ Kleinbaum / "Founders Kleinbaum and Brian Frezza"

- "grew out of Silicon Valley biotech startup Emerald Therapeutics"

- outside influence: “We looked at what other industries had done and asked what is different from DNA sequencing and biosynthesis that is different from other bio?”

- " With $2.5 million worth of equipment and 5,000 square feet of space, the whole operation now runs on Wolfram Language."

- easily reproducible, vs: "big pharma like Bayer can’t even get very good consistency"

- "now opening up its cloud lab to other research firms." (July 2014)

- "Founders Fund, PayPal co-founder Max Levchin and Schooner Capital supported the lab in the latest round of financing for a total of $13.5 million."

- "ontology-indexed database that will be accessible in the cloud" to allow "similar experiment" lookup

- "The goal, says Kleinbaum, “is to be the Amazon Web Services of wet lab experiments.”"

- "Outside researchers will be able to order up to 40 of the major in vitro molecular and cellular biology experiments using the robotic equipment at ECL. The plan is to expand to hundreds of other experiments in the next 18 months."

RS?: There are arguments about techcrunch in terms of RS. It looks like it's okay as long as 1) the article is written by a staff writer and 2) the author doesn't have a conflict of interest. This doesn't look like the case here.

- 40+ types of experiment; $3 million worth of lab equipment; equipment includes DNA synthesizers, advanced microscopes and mass spectrometers

- 15,000-square-foot facility

- DNAnexus provides cloud storage + analysis of DNA sequences, but ONLY the data

- Transcriptic provides an automated+controlled lab, but only for ~20 experiments

- average cost per experiment is $25

- "Having reproducible results isn't the only potential boon from labs in the cloud. There's also a financial advantage, since modern labs buy equipment costing anywhere from $100,000 to $240,000 per piece, which they use for perhaps 10 percent of their research."

- founded in 2010

- "DJ Kleinbaum, Emerald's other co-founder and co-CEO. Best friends since grade school, Kleinbaum holds a Ph.D. in organic chemistry from Stanford University; Frezza earned his Ph.D. in chemical biology from The Scripps Research Institute. "

RS?: CNET looks good for RS

Another possible (academic source) [2]

[edit]

- 89% of published papers contain methods available in a cloud lab, but only 3% are entirely composed of such methods

Emerald Therapeutics: Biotech Lab for Hire (bloomberg)

[edit]
July 1 was release of ECL web interface
"D.J. Kleinbaum and Brian Frezza, the company’s founders, grew up a couple houses apart on Emerald Drive in the suburbs of Philadelphia."
started as idea for "breakthrough antiviral therapy for diseases such as hepatitis and HIV." During this process, they developed tools to centralize control of disparate research devices and a database to consolidate information from every machine (rather than paper printouts or piecewise solutions)
backed by Peter Thiel's Founder's Fund for $13.5 million. This was the second round of funding.
In 4 years (to 2014), ran 400k experiments with about 1 million results, keeping track of machines and configurations to ensure reproducibility
early clients included Covert Systems Biology Lab at Stanford
praised by Sydney Brenner (nobel-winner), who said, “I foresee that the system could be used to repeat the published experiments in an objective way and allow companies in particular to assure themselves before committing large resources to development.”


Is it RS? I think so. In perennial sources, it's noted that Bloomberg itself is generally good, while its company profiles are usually not. I'm pretty sure this article was originally in Bloomberg Businessweek, since 1) that's where the picture seems to be from and 2) because the author seems to be one of the actual writers for bloomberg (not from ECL).


The Emerald Cloud Lab: Data Lessons Learned from Running a Remote-Controlled Lifesciences Laboratory (Wolfram Data Summit talk) [3]

[edit]

- (0:45) ECL started as an in-house tool to standardize experiments and experimental controls.

- (1:27) business case is SCALE

- (2:10) Describing experiments is really hard (ontology). Source: Amgen reproducibility problem. Why? Frezza argues that this is because describing experimental procedures is not done well.

- (4:03) That is, some important details in papers remain ambiguous, or have "intuitive language" --> PB+J analogy, non-machine-understandable instructions

- (6:08) In order to deal with this ambiguity, ECL had to define a new language to define what experiments look like in practice

- (7:04) Example: western blot. What are inputs? What are experiment details? (long list of options)

- (9:23) "Can control data from cradle to grave" -> solves problem of finding raw data, which is often difficult in the literature

- (10:41) Specifies all parameters for the machines, since that makes sense at scale. This allows additional specificity/clarity in the data

- (12:27) Everything is in the Wolfram Language on their platform

- (14:02) Final figures are linked all the way through the raw data and the input parameters, through the Wolfram Language-based setup

- (16:30) In scientific datasharing, lots of data is lost over time. Paper is not enough! Keeping a digital trail is important.

- (17:11) Example: Global warming data. The rising temperature graph doesn't include the metadata. Linked data is better because you can (digitally) add easily accessible datalinks

- (19:36) This linked data approach allows faster questioning of data and faster model-building


- article mostly about Stephen Wolfram and Wolfram Language

- "Brian Frezza, co-founder of Emerald Therapeutics, calls Wolfram Language 'a perfect fit for the science we’re doing' and said Mr. Wolfram’s design choices reflect a 'technological auteur' at work, giving his software a 'fanatical user devotion that you find in things like Apple products.'"

10.3389/fbioe.2019.00018

[edit]

- another RS

- one of two leading companies (with Transcriptic) towards centralized lab services through a software platform

- ECL focuses on a wide and flexible range of services

- ECL has a large array of analytical chem techniques

- ECL forcuses on fully detailed and reproducible methods

Jlevi (talk) 23:10, 29 October 2019 (UTC)[reply]

The Most Important Question To Ask New Employees

[edit]

- https://www.fastcompany.com/3035458/the-most-important-question-to-ask-new-employees

- article on Emerald Therapeutics

Jlevi (talk) 13:19, 23 February 2020 (UTC)[reply]

Jlevi (talk) 23:43, 22 November 2020 (UTC)[reply]

References

  1. ^ Buhr, Sarah. "Experimenting with Drugs in the Cloud". TechCrunch. Archived from the original on 11 July 2014. Retrieved 29 October 2019.
  2. ^ Jessop-Fabre, M. M.; Sonnenschein, N. (2019). "Improving Reproducibility in Synthetic Biology". Frontiers in Bioengineering and Biotechnology. 7: 18. doi:10.3389/fbioe.2019.00018. PMC 6378554. PMID 30805337.{{cite journal}}: CS1 maint: unflagged free DOI (link)
  3. ^ Frezza, Brian (4 September 2014). The Emerald Cloud Lab: Data Lessons Learned from Running a Remote-Controlled Lifesciences Laboratory. Washington, DC: Woflram.