Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Bringing spectroscopy into the COMCIFS fold

Hi, Peter,

thanks for your comments!

On 2017-02-23 11:25, Peter Murray-Rust wrote:
> I have been interested in getting chemical spectroscopists to adopt
> standards for publishing and sharing data for at least 20 years. In
> short: there needs to be communal will and there isn't enough to drive a
> useful process.

Well, at the moment we need to set up smooth data exchange for our SOLSA
project, so there is a need for common expandable spectroscopy standard;
also, other people are looking for similar things, as I have learned
recently.

The key point is that it needs to be simple -- we can not afford
spending much time to implement a complicated format/standard.

> There are standards - starting with JCAMP - which work
> for many potential uses. JCAMP - 30 years old - can do much of what is
> required for small to medium data sets and especially where there is a
> single technique.

> NIST / ASTM are creating ANIML https://www.animl.org/ which is about 10
> years old. I don't know what its trajectory is. It seemed bloated before
> any releases - e.g. it had 5 shells, including vendor.

Thanks, I'll have a look. Either we consider these standards, or take
parts from them to be compatible.

> Spectroscopy is more technically more varied and difficult than
> crystallography. There are proprietary instrument formats, deliberate
> vendor lock-in, complex multidiscipline experiments, data vendors, etc.

Well, vendor lock-in is something I am ready to fight against, and a
common open standard is worth having, if only for that reason.

> Unless there is a clear organisation with a future driving the process I
> think years could be spent getting nowhere.

IUCr?

> In short unless there is a strong spectroscopy champion I would not
> develop in advance of market pull.

We have several spectroscopy groups collaborating in our SOLSA project,
and they expressed genuine interest in having a standard for data
sharing. If we can come up with something simple enough, the investment
is worth doing -- and then we'll see.

I would say a standard is only as good as the software that supports it,
and only of there is considerable amount of available data in that
standard. If we can get these then the effort is worth it.

The worst thing that can happen is that people will start dumping
spectra in CSV files from Excel, without any standard or ontology
whatsoever. We would then risk loosing a lot of valuable measurements
just because they are not properly documented.

Regards,
Saulius

-- 
Saulius Gra┼żulis

VU Biotechnologijos institutas
Saul─Śtekio al. 7
LT-10257 Vilnius
Lietuva

Tel.: vidaus BTI:    4353
      Vilniaus BTI:  (8-5)-223-4353
      mobilus TELE2: (8-684)-49-802
      mobilus OMNIT: (8-614)-36-366


Reply to: [list | sender only]