October 26 28, 2011 D&A Meetings Proteomics Building, Rutgers

tribecagamosisΤεχνίτη Νοημοσύνη και Ρομποτική

8 Νοε 2013 (πριν από 3 χρόνια και 9 μήνες)

75 εμφανίσεις

October 26


28, 2011


D&A M
eeting
s Proteomics Building, Rutgers


October 26 Morning.

MQ/JW notes

Participants at RU:
J
asmine
Y
oung
, S
anchayita
S
en
, M
arina Z
huravleva,
C
henghua
S
hao
, S
ameer
V
alenkar, Tom Oldfield
, E
zra
P
eisach
, J
ohn
W
estbrook
, Z
ukang
F
eng
, Raul

Sala
,

M
artha
Q
uesada


by VTC:
L
uana
Rinaldi
, G
len van Ginkel
, Alex G
utmanas

Blue for
specific
further technical discussion

Red for
specific
further requirements analysis
/discussion


Agenda item:
D&A Dep mockups update review

a.

Tech impact on mock

ups


data clustering

b.

Ligand module components

c.

Sequence module components (reminder of the decisions made this summer)

1.

R
eview of the mock up updates for the DepUI.

a.

Use

a question mark (“?”) as a consistent icon for help information.

b.

Some additional det
ails need to be addressed for NMR & EM methods. Mockups for these
methods are being separately developed by Monica &
Aleks for NMR and Cathy and Arda
n
for EM.

c.

File upload page now has the context of the experimental method in order to present
appropriate upload options. Establishing the experimental contexts for X
-
ray, NMR, and
hybrid methods require further elaboration.

There needs to be coordinate input on method
obtained at the deposition distribution step/brain launch step.

d.

Need to natura
lly accommodate hybrid experiments (Aleks).

e.

Data harvesting vs file upload?
Incorporation of PDBextract functionality?
Harvesting
file
s
via PDBextract is supported by significant

infrastructure.
Will we e
nable PDBextract as a
module with a common interfa
ce look, rather than
integrating with the
middleware
?


f.

What will be required at upload?


Will we accept only a complete mmCIF as per the recent
format meeting.

a.

(Comment by Tom)
There is a timeline issue, that the new format will not be
expected
to role out in 2013


after the completion of the D&A system. We cannot
assume the new format will solve the D&A issues in the short term.

b.

(Comment by Tom)
We need to plan migration from PDB to mmCIF and to
encourage users to switch when mmCIF upload
becomes available.

g.

Highest le
vel questions presented at the B
rain,

ie. Experimental methods, what ID’s do you
want, then

first DepUI page would limit options based on the brain.

h.

First three pages of File upload must be aware of all methods/hybrids.

i.

Can
the questions
/suggestions

for the interface be encoded in Word or should there be a
image mockup of these questions? The group feels that the image mockups are important.

j.

The mechanics of uploading mandatory files needs further discussion. The current
mockup
provides separate UI sections for mandatory file types (e.g. model, experimental data).
Addressing multiple files of a particular type and optional data files needs further
consideration. It may be necessary to interpret the content of uploaded

data in order to
determine additional upload requirements.

k.

All data that must be uploaded is covered in the mandatory data requirements


those
items will be highlighted.
We need to identify auxiliary items.

l.

Replacement coordinates



after submission

-

h
ow is that uploaded and what are the
requirements for those cases?
Process and Requirements to be defined.

m.

Error text/diagnostics


what will the content and presentation look like for issues that arise
during file upload. Rules for accepting files


f
ile size, file type, parsing etc.

Reporting of
diagnostics from file uploads needs further definition for UI and middleware.

n.

Post submission upload requirements need to be elaborated further.

o.

Do all files go to the API for parsing prior upload.

Yes and No


JW elaborated

i.

Immediately obvious problem


file sanity check, all parts there? Format looks
good? Parsing issues? Etc


give immediate feedback to depositor

ii.

Down stream inconsistency


scientific issues that may require new file upload but
not recognized until latter in the process.

iii.

UI should indicate the validation that has been done.

2.

Polymer

a.

How much “missing data” will be filled in based on information that can be extracted from
the uploaded data ie. coordinates


Seqres?
Curre
nt system
s do not extract they
ask
depositors.

i.

Consensus
:
The source of sequence data should not be automatically populated
from the uploaded coordinate records. Sequence data can be parsed if it has been
properly integrated into the PDBx/mmCIF data file.

b.



S
amples


to be changed to macro
-
molecules.

c.

Cathy Lawson feedback


regarding the virus information box


move to a symmetry page
that provides the crystal system, NCS.
Concurrent conversations have occurred at the EBI


so we have consensus on this.

d.

Electronic mock up reviewed


File upload for matrix to be added.

e.

“The structure contains”
-
> “Macromolecule is part of a” (e.g. chimera antibiotic/inhibitor
complex). These options will flag the collection of additional data items and will be used to
subsequence processing. The role of each feature needs to clarified in terms of how it will
be used by the D&A system. The presentation of current options may not be sufficiently
clear (Chenghua).

f.

H
armonize terminology, add non of the above, virus move
s to separate page

g.

P
urpose of this information is to steer deposition to additional specific pages


now or in the
future. Requirements should capture the current intent.

h.

Consensus reached on removing “tag”

from the UI for
m

i.

Fragment decomposition of
polymer entities needs to be added to the data dictionary
(Jasmine).

j.

UI organization for collecting chimeric input can be done as an extension of macromolecule
entry page. An additional set of forms is not needed (Glen demo).

k.

Additional UI requirements
sh
ould
be captured by updating printe
d pages from working UI
where available rather than working on the corresponding excel mock ups.

l.

UI implements a field copy/”add another” option between forms. This can be used to
simplify input of sequence and source

information.

m.

Author identification remains an open issue.
Title and Author


how do we manage consistency
with historical records?

Don't over engineer relative to the quality of data that can be captured.
In the future “research contributor identificati
on”
-

need WW social security #...

n.

Group should review revised UI demonstration site at the beginning of the week for
discussion at the next group meeting on Nov 3.

o.

Under Contact information, for fragment information there is a cross dependency between
source and sequence. “Add fragment” option needed.


i.

“Related Entries” on the Admin page needs to include additional input form elements for
database code and database name. Form elements also need to be provided for the
nature of the relationship need an
d the additional details.

p.

Glen will have all of the electronic mock ups in place by Oct 31 for general review and comment.
Nov 3 ProjTM mtg agenda item will be a review of the pages.

i.

Initial performance assessment
-

HTTP vs HTTPS pages should have simila
r
performance.

ii.

Sample

Macromolecule page is the most critical to be developed for next week.

iii.

Ligand page still required more design work so hold off on that one.

q.

When will PISA calculate the biological assembly?
When the depositor triggers validation?
Do
all calculations at once?

3.

Release Status



a.

“Select DB name form a pull down list”

need to redesign this feature.
Per ID there are two
more items, one drop down, one text box.
GV will address.

b.

Supersede option


when do we ask this? Critical to releas
e not in the process for initial
deposition. Will stay with current design.





Comment by Tom


The MW has 4 release stages


see timeline

The MW component development effects the functionality of the depUI and its use with “real” data.
Until release 1

is complete the de
pUI will only display hard
-
wired example data. After release 1, the
depUI will display
“real” data from an mmCIF


but it will be read only. It can be changed, but updates
in the UI will NOT be put back to the data model.

The December

22
nd

release (release 3) of the MW will
contain all the read/write functionality to and from a mmCIF data model file.

Release 4 (Feb
-
2012) will
contain all the “expected” functionality, in particular, the auto
-
complete functionality.