HCS Platform comparisons

[Kate Luby-Phelps]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello everyone,

I am interested in discussing the relative merits of high throughput HCS
platforms based on your experiences. Specifically, we are considering the
Perkin Elmer Opera, the Yokogawa CV7000 and the Cellomics VTI. We are
particularly interested in knowing:

1)  the actual time to image a 1536 well plate in 4 channels

2) the ease of porting the raw data to third party software packages

3)  the steepness of the learning curve

4) the reliability for each instrument and quality of customer support

 We are not necessarily interested in the analysis packages that come with the
instrument but feel free to offer comments on that as well.

Thanks!

Kate Luby-Phelps

[Laszlo Komuves]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Dr. Luby-Phelps,

You might also want to consider the GE InCell
6000<http://www.gelifesciences.com/aptrix/upp01077.nsf/content/Products?OpenDocument&parentid=658508&moduleid=168006>which
is very fast (actually it was faster than the Opera in side-by-side
comparison during our evaluation).
No commercial interest but I am trying to get one for our group.
Best, Laszlo



On Tue, Oct 11, 2011 at 11:22 AM, Kate Luby-Phelps <
[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hello everyone,
>
> I am interested in discussing the relative merits of high throughput HCS
> platforms based on your experiences. Specifically, we are considering the
> Perkin Elmer Opera, the Yokogawa CV7000 and the Cellomics VTI. We are
> particularly interested in knowing:
>
> 1)  the actual time to image a 1536 well plate in 4 channels
>
> 2) the ease of porting the raw data to third party software packages
>
> 3)  the steepness of the learning curve
>
> 4) the reliability for each instrument and quality of customer support
>
>  We are not necessarily interested in the analysis packages that come with
> the
> instrument but feel free to offer comments on that as well.
>
> Thanks!
>
> Kate Luby-Phelps
>

--

László G. Kömüves, PhD, Senior Scientific Manager

Center for Advanced Light Microscopy (CALM) | Department of Pathology –
Genentech, Inc. | 1 DNA Way, South San Francisco CA 94080-4990

E-mail: [hidden email] | Phone: [phone]

[mahogny]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

>From what I have seen, these systems have the big disadvantage of being
very locked to (and limited by) the software that comes with them;
they're even more of a black box than "normal" microscopes for which we
already have to fight to get access to the hardware we have paid for.

In other words, the chance that you are paying for a very overpriced
paperweight is much elevated. I would not buy one.

With that said, what are other people using to feed multiwell plates to
their microscopes? Just out of interest.

/Johan

> > Hello everyone,
> >
> > I am interested in discussing the relative merits of high throughput HCS
> > platforms based on your experiences. Specifically, we are considering the
> > Perkin Elmer Opera, the Yokogawa CV7000 and the Cellomics VTI. We are
> > particularly interested in knowing:
> >
> > 1)  the actual time to image a 1536 well plate in 4 channels
> >
> > 2) the ease of porting the raw data to third party software packages
> >
> > 3)  the steepness of the learning curve
> >
> > 4) the reliability for each instrument and quality of customer support
> >
> >  We are not necessarily interested in the analysis packages that come with
> > the
> > instrument but feel free to offer comments on that as well.
> >
> > Thanks!
> >
> > Kate Luby-Phelps
> >
>
>
>

[Emmanuel Gustin]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello Kate,

My two cents:


1) Very difficult to estimate in general, because you nearly always have
to acquire multiple fields per well. Manufacturers will usually give an
estimate for 1 field per well, which isn't very realistic. Assuming you
have a reasonably homogeneous cell population, you want at least ~ 200
cells per well for acceptable statistics in a high-content screen, and
that means that the number of imaged fields per well strongly depends on
the selected magnification, the size of the cells, and the degree of
confluence. In my experience, taking one field per well is fairly rare,
and 3 to 6 more common. (But the large-area sCMOS cameras on the latest
generation of HCS readers have an advantage here.)
Autofocus time is also a significant factor, and influenced by the
objective and by the plate quality. On the Opera I estimate that with
the 20x water immersion objective, stage travel time and autofocus time
add up to about 800ms per field; perhaps a bit less in 1536, and a bit
more in 96... Add the exposure time, and multiply by the number of
fields and wells.
And exposure time is also an "it depends". The Opera and the high-end
Yokogawa benefit from having multiple cameras, gaining time by combining
several image channels in a single exposure: In theory, you can do 4
channels simultaneously on the Opera, if you include the non-confocal UV
channel. In practice, the crosstalk and signal ratios of your sample do
not always allow such multiplexing, and then you may have to do multiple
exposures with different lasers and cameras; say 2 exposures for 4
channels.


2) Not too bad, despite the much-decried lack of a standard for HCS
data. Most readers produce images in some variation on TIFF, and numeric
and metadata in tab-delimited text or XML files. The lack of a standard
mostly makes the conversion of the data sets very tedious, especially if
you have multiple types of reader. The BioFormats library makes it a lot
easier to read and process the images, however.
The Opera, for example, produces .flex image files that are really
multi-page TIFF with an extra scaling factor and an XML metadata block.
(Avoid the LuraWave compression option.) A recent software release
changed that file format to single-page TIFF, which most users find
easier to process, but we don't have that upgrade yet. The measurement
and analysis files are written as XML files.
External software may stumble on the dimensionality of the data sets,
however. Not just the XYZT dimensions of the collected images, but the
experimental reality of multiple fields per well, multiple scans of a
plate, multiple image analysis runs of a scan, ...
Another factor: Network, server and storage configurations need to be
discussed with your IT department. The Opera comes standard with a set
of servers (which may be put in a rack under the instrument), and it
needs gigabit network to your network storage (somewhere from 5TB
upwards). Images are stored by the servers, not by the instrument.


3) As far as data acquisition is concerned, I think most of the
instruments are a bit easier to use than a standard confocal microscope,
having fewer options to select from. It still requires training and some
insight in microscopy. We find a good lab technician can run the Opera
routinely, but a microscopy specialist may have to assist with
optimizing the exposure and protocol settings. (Efficient use of the HCS
instrument also requires a fast feedback loop between the image/data
analyst and the people who develop the assay and acquire the images.)
Image and data analysis are more challenging. There is a strong
trade-off between the ease of use and the flexibility and power of the
package, and all vendors are struggling with it. The Opera, for example,
used to be delivered with a package that was very flexible but hard to
learn for non-programmers. The latest software upgrade added the much
easier user interface of the Harmony software. Other vendors have gone
the other route, starting out with simple-to-use but inflexible packages
and adding more power in later releases...


4) Most HCS readers are inherently complex instruments, and that has the
usual consequences for their reliability. The Opera is not too different
in this regard from a motorized confocal microscope, you can have the
usual problems with lasers, beam coupling, and fine mechanics. Find a
location with a stable temperature and free from vibration, and ensure
users take care to avoid damaging the objectives or the focus mechanism.
Our reliability and level of support have been good. As always, allow
for the reality that you have to call on a limited number of technical
specialists and you may have to wait for a week or so before one can
come on-site...


best Regards,

Emmanuel




-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]]
On Behalf Of Kate Luby-Phelps
Sent: dinsdag 11 oktober 2011 20:22
To: [hidden email]
Subject: HCS Platform comparisons

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello everyone,

I am interested in discussing the relative merits of high throughput HCS

platforms based on your experiences. Specifically, we are considering
the
Perkin Elmer Opera, the Yokogawa CV7000 and the Cellomics VTI. We are
particularly interested in knowing:

1)  the actual time to image a 1536 well plate in 4 channels

2) the ease of porting the raw data to third party software packages

3)  the steepness of the learning curve

4) the reliability for each instrument and quality of customer support

 We are not necessarily interested in the analysis packages that come
with the
instrument but feel free to offer comments on that as well.

Thanks!

Kate Luby-Phelps

[John Bradley]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Dr. Luby-Phelps,

Just a couple of comments (I'm just a user - no commercial interests):

We use both the ImageXpress Micro and ImageXpress Ultra. And we run analysis
using Molecular Devices's PowerCore configuration. The analysis modules are
user friendly and PowerCore can be exceptionally fast. Since installation (2
yrs), we've had few problems with the systems and their tech support has
been great.

We recently demo'd GE's IN cell 6000 - the software looked very impressive
and very user friendly. They have licensed Spotfire and integrated it into
their acquisition software.

Cheers, John



On Tue, Oct 11, 2011 at 2:22 PM, Kate Luby-Phelps <
[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hello everyone,
>
> I am interested in discussing the relative merits of high throughput HCS
> platforms based on your experiences. Specifically, we are considering the
> Perkin Elmer Opera, the Yokogawa CV7000 and the Cellomics VTI. We are
> particularly interested in knowing:
>
> 1)  the actual time to image a 1536 well plate in 4 channels
>
> 2) the ease of porting the raw data to third party software packages
>
> 3)  the steepness of the learning curve
>
> 4) the reliability for each instrument and quality of customer support
>
>  We are not necessarily interested in the analysis packages that come with
> the
> instrument but feel free to offer comments on that as well.
>
> Thanks!
>
> Kate Luby-Phelps
>