HI Michael,
You are right, it was a typo - Factors are Biomaterial or Population characteristics and not related to assays.
I really like the way you are associating the protocol to the transition step, makes a lot of sense. Sometimes we don't have that level of detail captured for the experiment; hence we are associating it with the design step as that is the step that produces the protocol and the rest of the steps follow it.
Best,
Sudeshna
On Nov 30, 2011, at 11:53 AM, Michael Miller wrote:
> hi tim,
>
> thanks for your comments, i'll be interested to see how the Factors are related to the assays, not sure how subclasses would work since one actually needs different factor values for the factors, i.e. a time factor can't tell you where in the time sequence the assay was done.
>
> interesting about the protocols. in implementing Annotare, the fged sponsored open source tool for creating and editing MAGE-TAB documents, we faced this same issue. for various columns, like characteristics, species and so on, we offer drop down selections based on ontology views from NCBO. when it came to protocols, you're exactly right, one has to consider the position of the column in the mage-tab SDRF file in order to know what type of protocol to use to populate the drop down. i'm not even sure if our UI guy has quite got it right yet. but in the BioRDF use case where we are looking at translating MAGE-TAB to RDF triples, since it is already known, creating the triple associating the instance of a protocol to the transition between an instance of material and/or data to the next instance of material and or/data was trivial. and if one is doing certain investigations where it's the protocol that is being varied for a certain step, this would be a tremendous loss to understanding/replicating of the results to not have the link.
>
> the other interesting case which i haven't seen anyone solve yet is the fact that factors are duplications of characteristics of the samples or protocol parameters but it is difficult to model that relationship. i've thought that might be some sort of 'is_a' relationship.
>
> cheers,
> michael
>
> From: Tim Clark [mailto:tim_clark@harvard.edu]
> Sent: Wednesday, November 30, 2011 8:30 AM
> To: Michael Miller
> Cc: HCLS IG; scidiscourse@googlegroups.com; Alberto Accomazzi; Bradley P (ELS-SDG) Allen; Sophia Ananiadou; Philip Bourne; Gully Burns; Ronald (ELS-SDG) Daniel; Rahul Dave; Alf Eaton; Alexander Garcia Castro; Matthew Gamble; Yolanda Gil; Alyssa Goodman; Paul Groth; Tudor Groza; ellen hays; Maryann Martone; David R Newman; Antony (ELS-CAM) Scerri; Jack Park; Silvio Peroni; Steve Pettifer; Philippe Rocca-Serra; Cartic Ramakrishnan; Jodi Schneider; David Shotton; Kaitlin Thaney; Karin Verspoor; Lynette Hirschman; Susanna-Assunta Sansone; Kees van Bochove; Katy Wolstencroft; Jun Zhao; Taylor, Michael (ELS-OXF)
> Subject: Re: SWAN-myExp-OBI-ISA
>
> Hi Michael,
>
> Thanks for your comments. I appreciate them. I've incorporated the one about factors into a subsequent version of the model, and decided to pass on the protocol relationships to data output, for pragmatic reasons, although I am certain you are formally correct.
> Factors were associated with the assays via a series of subclasses in previous versions of the model. In this model I omitted the subclasses for improved readability but I should not have omitted the relationship with assays. I am adding them back at your suggestion.
> Protocols we decided to leave formally unconnected to the data output in this model for simplicity. As my colleague Sudeshna Das points out, Protocols "... could be assay protocols, sample processing protocols or data processing protocols." It is always a tradeoff and judgement call between increased formalization and additional burdens on the scientist to input the data.
> Best
>
> Tim
>
> On Nov 28, 2011, at 2:05 PM, Michael Miller wrote:
>
>
> hi tim,
>
> looks like you all are making good progress.
>
> this is just from looking at the diagram, so this may already be covered,
> but both protocol and factor are also related to other concepts.
>
> protocols would be associated with 'has_output' and 'computed_from', and
> there would be factor values associated with each factor of the study and
> each assay would be associated with a set of factor values, one pre factor,
> e.g. for dosage and time factors an assay might be associated with 1mg and
> 10 seconds.
>
> also, is there any desire to track sample processing?
>
> cheers,
> michael
>
> Michael Miller
> Software Engineer
> Institute for Systems Biology
>
>
> -----Original Message-----
> From: Tim Clark [mailto:tim_clark@harvard.edu]
> Sent: Monday, November 21, 2011 7:32 AM
> To: Tim Clark; HCLS IG; scidiscourse@googlegroups.com; Alberto
> Accomazzi; Bradley P (ELS-SDG) Allen; Sophia Ananiadou; Philip Bourne;
> Gully Burns; Ronald (ELS-SDG) Daniel; Rahul Dave; Alf Eaton; Alexander
> Garcia Castro; Matthew Gamble; Yolanda Gil; Alyssa Goodman; Paul Groth;
> Tudor Groza; ellen hays; Maryann Martone; David R Newman; Antony (ELS-
> CAM) Scerri; Jack Park; Silvio Peroni; Steve Pettifer; Philippe Rocca-
> Serra; Cartic Ramakrishnan; Jodi Schneider; David Shotton; Kaitlin
> Thaney; Karin Verspoor; Lynette Hirschman; Susanna-Assunta Sansone;
> Kees van Bochove; Katy Wolstencroft; Jun Zhao; Taylor, Michael (ELS-
> OXF)
> Subject: SWAN-myExp-OBI-ISA
>
>
>
> The information in this e-mail is intended only for the person to whom
> it is
> addressed. If you believe this e-mail was sent to you in error and the
> e-mail
> contains patient information, please contact the Partners Compliance
> HelpLine at
> http://www.partners.org/complianceline . If the e-mail was sent to you
> in error
> but does not contain patient information, please contact the sender and
> properly
> dispose of the e-mail.
>
>