4:00 - 4:05pm: Announcements and Approval of 2/4/21 Minutes (5 minutes)

 

David MacAlpine:

  • Calls meeting to order and reminds individuals to sign the sign-in sheet and pick up any agenda should they need one. 
  • Motion to Approve the minutes of Feb 4th 2021 – APPROVED
  • One other announcement from Ryn Nasser

 

Ryn Nasser - The new ITAC website was just launched with Sites@Duke Pro system that was on Tuesday.  We are very excited about it she’s working on migrating the minutes over now.  If there are any other minutes from 2013 please let her know.  The goal is to have all the agendas and minutes up in the next couple of months.

 

4:05 - 4:50pm: Special Guest: Jenny Lodge  (15-minute presentation, 30-minute discussion)

  • What it is: Each year ITAC has the opportunity to hear from selected senior leaders. This week we are pleased to welcome Duke’s new Vice President for Research & Innovation, Jenny Lodge, who will answer questions and share her perspective as it relates to the intersection of technology and research at Duke.
  • Why it’s relevant: As the university’s chief research and innovation officer, Lodge leads oversight of Duke’s $1.2 billion annual research portfolio, including grants administration, ethical practices, and commercialization. She works with campus and medical center research staff, faculty, and trainees, and is a key figure in Duke’s connection with external partners. ITAC members are invited to pose their questions to her.

 

Minutes:

 

David Macalpine :

Welcomes Dr. Jenny Lodge as the new Vice President of Research & Innovation overseeing all of Duke’s research portfolio both on the medical and university sides, “Big Duke” if you will.  She is coming from WashU where she was the Vice Chancellor of Research and Senior Associate Dean of Research at the School of Medicine.

In addition to her role as Vice President at Duke,  she is also Professor in the MGM, Molecular Genetics and Microbiology, space  and has an active research lab there. So, she literally has boots on the ground and is familiar with running an active research group at Duke.

 

 

David MacAlpine – Now that you have had almost a year to be here, what were your initial reactions when you came about, and what similarities and dissimilates between Duke and WashU?   Are there any unique challenges and opportunities that you are excited about at being at Duke?

 

Jenny Lodge - I’ve enjoyed my transition to Duke.  I’m finding it to be an incredibly exciting place.  Seems to be a very interdisciplinary and cooperative kind of research here.  It really feels like a special place, people are enthusiastic about being at Duke and the time of work they can get done here. In terms of the differences and similarities between Duke and  WashU. 

 

There is a very strong focus both at WashU and Duke for the undergraduate experience making sure that the undergrads have a lot of support .  Both have strong graduate programs.  Really are a lot of similarities..

 

There are some differences as well especially when it comes to the IT space. WashU has been reevaluating how to organize its IT.  

 

Duke:

You have the campus the side IT and the Health System IT, and there are clearly barriers – it’s hard to transfer data or you can be on one system but it’s hard to navigate between systems. There are good reasons for that data with protected data – health stuff you have to be super secure penalties for data loss are expensive. 

 

WashU:

There were two divides between the CFU – Central Fiscal Unit schools (campus) versus the Medical School … The medical school had a lot more security, what type of equipment people could use, and how and where the data was stored or transmitted.   And there were some problems getting things across to campus. Second divide was between the Medical School and the Health System. The Health System was an independent entity not owned by the University.  There was a lot of effort to get research data into one place, it took a long time to figure that out for the Medical System and the University to get to a happy place to get that setup.

 

So, at any place where you have a health system with clinical data and a university with physicists, historians, and chemists, there is a difference in culture and a difference in security. And trying to figure that out is a challenging and one of the things Tracy has done which I think is a strength here is taking a deep dive on this needs assessment and trying to figure out what the next steps should be and how research should be supported and the next steps to move forward. 

 

Having this kind of group that includes students and faculty and individuals from IT to discuss these matters to have discussions and analyses with high-performance computing, storage, addressing near-term and long-term and moving into Archives.  

 

Jenny Lodge:

Introduces Chris Freel as  Assoc Vice President of Research overseeing the Campus side, in her office thinking of research concerns across campus.

 

Chris Freel – The effort Tracy has led with this group over the last year it has been really good to get the report as it is because it’s highlighted by various groups all across campus.

 

Sunshine Hillygus  - Political science.   I am relatively new to ITAC, one of the things is I was a squeaky wheel in meetings and so I got an invite to join ITAC.  So, this self-study that has been done about IT but it more than IT, so the support needs that have been identified in this very comprehensive study are not only under IT?  What is the next step, since it is not just within IT?

 

Jenny Lodge– First of All Tracy has been sharing this with me and to the Provost to raise it at a higher level. She’s also on the agenda for UPC.

 

Tracy Futhey – I will bring this to UPC – University Priorities Committee – which is a group that helps to vet the expressed needs of new priorities, we all know these needs come with $$ so they can’t wave a magic wand but they can be on the same page to raise the priority.  So, I speak to them next week.  Then we will be talking with Deans.    We also have a new University Librarian – to help with this he has been here about 6 weeks but we also will need to bring him in on how to marshal all of our resources.  

 

Jenny Lodge– Engaging the Health System is absolutely critically important as well. So, I report jointly to Provost Sally Kornbluth and Gene Washington (Chancellor of the Health system )

 

Tracy Futhey– Gary Hoke who has joined us as well, is responsible for the research technology in the School of Medicine and he reports to DHTS.  He has a foothold in all three domains, being in this meeting, being in SoM, and being in DHTS.  

 

Jenny Lodge –  What I hear you asking is, is there an organizational structure to report this up through, well Vince is at the top of the organization.  But I think  UPC – with Daniel Ennis and Sally, and Gene this is important and they are the ones to move this forward.  One of the things I was really impressed with when I interviewed for this position I liked how aligned the leadership was on their vision of what Duke needs to do to move forward.   We have a really collaborative team to work with, I’m not worried about Tracy and Librarian and I, working together.  The Health System will have very different priorities and we have to respect that and they respect ours.

 

Tracy Futhey– I think we have also gotten an agreement to offer and provide the right set of services and support on behalf of all the faculty and not worry about whose empire it’s in, who owns what who does what.   What works for Duke and if we can keep focused on that we will be better off.

 

Jenny Lodge – an example is two positions that Chris has posted…

 

Chris Freel – one is an Executive Director in the Protected Data Space as well as a Program Director that is more tactical, boots on the ground.   This is shared position between OIT and the Research space and is the first of its kind, which is a direction we are now seeing in other Higher Ed spaces.

 

Terry Oas – Department of Biochemistry – One of the basic science departments. You have accurately perceived the barrier that exists in both actual and literal data storage and transfer, there is also a barrier that is policy related.  A big source of that barrier is administrative, administrative decisions not intrinsically related to activities that have to be recognized.    So, the health system has to be firewalled and treated very differently than our non-health system.  But I don’t consider any of our basic sciences departments fit into that except Pathology which is half of the basic science and half clinical.  None of the other basic sciences fit into the description of a health delivery system.   They don’t have any PHI data or rather not any more than the little that a campus side department has. 

 

Yet over ~15 years, we have been dealing with the policy set by the SoM or priorities of the Health System   So if OIT has a different policy and that has caused a problem… because if you are in SoM and you are in basic sciences you are held to SoM, not OIT. Do you think your office could/would, should impact policy decisions for the administration and attituded of upper administration of SoM regarding basic sciences support and use of OIT services?  Can your office actually change those attitudes?

 

Jenny Lodge – I would only be able to influence, but I don’t have the authority to change the policy of SoM.  Individual schools have the authority.  

 

Terry Oas - I’ve been talking about this for more than a decade.  OIT has made great strides to improve the network to accommodate our needs and is much more configurable, however at the time it was a point of money but not so much,  it’s now less about money and technology but it’s more policy related. 

 

Tracy Futhey - Point well taken, and the report tries to advocate as recommendations around some of these concerns is that we think about a holistic assessment and more of a risk-based determination of the level of security or the implementation of the polices. As you said Clinical is closer to the Health system, and Basic sciences are closer to campus, but there is a variation and a spectrum across there but we have to find what fits for the use cases.

 

Terry Oas -  Is it true that at this point physical location is less important in deciding if someone with PHI based system is on a particular network on campus.

 

David MacAlpine - One of the unique aspects of Duke is we are all integrated with the undergraduates & graduates on both campuses on the same networks with a mixture of populations. Many of us have labs with a mix of all these students.

 

John Board - I know you went to a training about how to keep your research grants…  And you have to swear fidelity to MyResearchWebsite in order to keep your grant.  So also, on that website, it says “We have 46 offices to help you do your research.”  That is your website…

 

Jenny Lodge - Training is something that we need to be looking at more carefully. As I’ve gone through the RCR training and another one, I do feel like the training gets information added because people feel like it will be useful.  It’s nice to know that the resources are there, but I don’t know that you need to be trained in it.

 

John Board - Why do we have 44 offices trying to tell us how to do our research when we should have 1? Under the current system, it is up to faculty to magically determine which of the 44 offices may and do give conflicting guidance.  You get slapped on the wrist after the fact for failing to go to office #37. This is part of the larger global frustration.  Some of this is technology workflow, and most of this is policy.

 

Jenny Lodge - So Chris and I have been looking at the org chart and see that some of the offices have duplication of effort or have grown up independently.   You don’t want the IACUC office –Institutional Animal Care and Use Committee, in with the IRB and Contracts. You need to have separate offices to handle those items.

 

John Board - Faculty shouldn’t care about that division which is

 

Jenny Lodge -  One of the offices that is growing is DORI – Duke Office of Research Initiatives – they are research navigators which are  live human beings, pick up the phone, and or email and they will get back to you pretty quick, to help you figure out who you need to talk to for what rather than relying only on My Research Home which some find a little overwhelming because it has so much information on it.

 

Tracy Futhey asked a question regarding myreserachhome on behalf of Stefan Bass, who is out of the country.

 

Sunshine Hillygus -  Regarding much of our research administration support and systems, I was going to emphasize that this didn’t come from IT this came from the medical side.  There has been more recognition that we need to get more faculty at the table rather than just replicating medical systems on the campus side. Previously we have given feedback and held focus groups, but when systems were already built it was too late.  The same thing will happen the next time if we still have the organization structure where the campus side is told the medical side has adopted this and it’s more cost-efficient if the campus faculty simply adopt an existing system.   Is there any appetite for revisiting that structure and that fundamental problem?

 

Jenny Lodge – You’ve described they way things have been implemented, in part because of prior problems that were experienced at Duke and a need to move fairly fast based on feedback from our federal funders.

 

Jenny Lodge -  One of the things Tracy and I both do is sit on committees that look at these various initiatives.  For example the PDC, and it will go away on June 30th, 2023.  And those faculty are going to be integrated in SoM, and they need an effort reporting system, they need to be able to make this work for those faculty. And they have to do it in a pretty fast timeline for implementing an IT solution for this.   Tracy and I and Sally are saying that’s great for SoM but we are not going to roll this out to the rest of the campus.  The campus has completely different needs in this area.  One big difference is people having 9-month assignments vs. 12-month assignment.  A lot of the schools on the campus side have 9-month appointments, and I know that is not in their heads so we are not going to implement that across the campus.

 

Robert Wolpert -  A related question a few years ago the Provost to be agreed to commit some money to cover the technology needs of some faculty but then time goes on and the needs evolved.  At some point faculty input needs to be involved in that process.  I wonder if you have any ideas on how to do that?

 

Jenny Lodge -  So it seems that this group, ITAC,  is a pretty good group for getting that input.  I hate to get more committees involved when this could be the right group already. 

 

Tracy Futhey -  How are you thinking about bringing in more faculty voices into this process?

 

Jenny Lodge - So we have a couple of committees that give out office advice.  We get the deans and vice deans together.  There is also EROC – Executive Research Oversight Committee – whch includes university leadership and a few faculty representatives and we recently expanded the roster to include additional research-active faculty to give their input.  We want to make sure they feel empowered to bring input to the process.

 

Jenny Lodge - Another big piece of this is a new Assoc Vice President in my office for the Social Sciences, Behavioral Sciences, Humanities, and the Arts  - so we need someone in that space to represent that space.   Our office is heavily Engineering and Bio-Medical based, and we need someone that is that way by identity to think of things from that perspective.

 

Colin Rundel –  from Statistical Science – So our department deals with data that runs the gamut, whether to totally public and nobody cares about it all the way to PHI.  So it’s one of the things where we fall in the gap between technology and policy.  So the solutions that exist or the policies don’t support it. The PRDN (Protected Research Data Network) is a disaster because it is really difficult to work it.  In some instances, it works great but for a large part of it, it’s not relevant because the data doesn’t warrant being mandated to the policies of the PRDN. We would much rather have guidance knowing that the data is protected and on a shared system but doesn’t need to be on PRDN and as such here are the requirements hopefully your IT person can get it together.

 

Jenny Lodge – I agree, deferring to Chris Freel – That the executive director of Protected Research Strategy will come up with solutions that right sizes the environment that don’t overprotect the data / processes so that it doesn’t make it difficult to manage.

 

Chris Freel – There are many things that don’t need that level of oversight…

 

Jenny Lodge – A valuable matrix was created over at the med school for people to evaluate the security risk of their data…it helped them decide on how to store the data between on campus or cloud.

 

Colin Rundel  – It needs faculty voices – but we have dealt with lawyers and they are safest answer is saying NO but that is not always the answer.

 

Jenny Lodge  – So we have to make the solution appropriate to the level of  risk. And we need to make adjustments… based on the risk.  It’s going to take some time

 

David MacAlpine –  We have had previews of Duke’s Data Policy about what will take effect in the next year.  Are there any updates?

 

Jenny Lodge– A year ago  there was an request for comments… It’s been an iterative process in the policy.

 

The stage that it is at is a draft policy and there is support needed to help the faculty know how to adher to the policy. The policy is going ECAC – Executive Committee’s Academic Council.  We have talked to ECAC and the current plan is to release the policy it in Nov/December, but delay implementation until May so people have time to acclimate to the policy and understand what’s in it.

 

Jenny Lodge – The other thing that changed in the draft is that it clarifies that data that originates outside of Duke is not considered to be owned by Duke…

 

Tracy Futhey – but we can share our data…?

 

Tim McGeary – That was intended but not clearly written  - now it is written in there.

 

 

 

4:50 – 5:15pm: Update on Research Computing Needs Assessment – 25-minute discussion

  • What it is: ITAC spent most of the summer hearing first-hand accounts from seven faculty groups about challenges and opportunities with research computing, broadly defined, and how computing tools and expertise might best support work throughout all the disciplines represented at Duke.  Our colleagues identified numerous issues with Duke’s current personnel, processes, and technology supporting research computation.  Today, Tracy and others on her senior leadership team will lead a discussion of the draft written report which attempts to summarize the current state of Duke’s research computing needs.
  • Why it’s relevant: When finalized, this report will be circulated to senior leadership and beyond to lay the foundation for the subsequent effort to make and cost specific recommendations for improvements in Duke’s research computing support.  ITAC will be extensively involved as this process continues throughout the fall semester.  

 

 

Minutes:

 

Tracy Futhey –I have circulated the draft report including 13 pages of deep gory details to the faculty of ITAC, it’s not circulated broadly, once the working group and faculty agree we will circulate it further.  So, this is an overview of that draft and how we’ve gotten here…

  • 40 Faculty and various support personnel were involved in the discovery process

this led to

  • 6 major findings and 10 recommendations in 3 Thematic areas:

 

  1. People - We really don’t have a sufficient set of personnel supports – a person with one foot in IT and one foot in my domain.

Recommendations:

  1. We need a service could support those research needs. The report doesn’t prescribe where these people would be housed organizationally or whether they would be technical staff, academic staff, librarians, post docs or otherwise. We need these “domain IT experts” everywhere, engineers, social scientists, humanities, basic sciences… this need was clearly expressed by every group of faculty.
  2. We also need training programs for faculty, research staff, graduate students, and undergraduates who support and engage in research.

 

  1. Process - We have great infrastructures and organizations that seek to collaborate.  However, we have faculty that are frustrated they find it difficult to navigate and organize things.

Recommendations:

  1. We need an external review to review and weigh in on the university and health system IT alignment related to research support.  This doesn’t presume an outcome but it recommends a process for an external review.

 

  1. Process & Conflicting Compliance & Security Measures for Campus and SoM

They are not as risk based as we would like.

Recommendations:

  1. How do we launch an evaluation of some of those policies that have been asked? How do we gradate the protections there, “without compromising regulatory compliance or the protection against sensitive data loss”?

 

  1. Duke Compute Cluster - They were valuable and good, but were not as expansive or flexible as our researchers need

Recommendations:

  1. Look at cost and funding to see it could be expanded.
  2. How we make our offerings a bit more flexible, a bit more supportive of data collection, statistical analysis machine learning, raw compute power etc.

 

  1. Computing Sherpa or Concierge Service -Researchers don’t know enough about what’s available to support them, at Duke or in the cloud… 

Recommendations?

  1. Having a computer Sherpa or Concierge – so they can help guide you with what you need to do and where to go, as an ongoing service.

 

 

  1. Research Data Storage - Difficulty of storing data over the lifecycle of a research project, including archiving concerns, affordability, and difficulty navigating. Not to mention risks when investigators are writing scripts themselves and then having to hold their breath to execute the movement data across platforms.

            Recommendations:

  1.       Implement new storage  options..
  1. Approach new long-term solutions as an eco-system. Recognize differences in storage needs across research lifecycle and automating to the greatest degree possible.
  2. Data Access and Data Use and how to manage and license data sets 

 

 

So, Where are We Now:

 

We are at the point to submit our 6-page report that faculty and the working group have seen and 35+ pages worth of Appendices – in pretty gory detail document the process and experience what helped us get there.  It’s long and painful, OIT is a steward of the data.

 

Where Next :

  • Going to UPC – University Priorities Committee
  • Do the findings / recommendation ring true to this group?
  • Are further steps need to secure faculty support for the report?
  • Anything further needed to capture research needs elsewhere> without delaying release of the current report.

 

Robert Wolpert – Statistical Science  - There is a thought to try to hire the IT area specialist – Right now we can’t recruit and retain a system administrator, without the additional constraints.   So, the recommendation – is build these guys…

 

John Board – Maybe we can get the folks to be grown internally, build them rather than hire.

 

Robert Wolpert – You can use OIT’s history as evidence that works…

 

Terry Oas – You can use OIT’s history as evidence of how that works.  Your turnover rate is quite low, you have built your staff in a significant way.

 

Tracy Futhey - Our OIT staff includes a few people with Ph.D.’s but we don’t have a feeder system that is based on post-docs etc.

 

Terry Oas – Your process for recruiting and retaining – could be used though as an example

 

John Board – Sometimes it might be a post-doc or it could be a bright undergrad like Zoey

 

Tracy Futhey – The needs will have to be attuned to the specific domain, a social scientist needs is totally different to basic science needs.

 

Terry Oas – The reason why I bring up retention is because the person you are describing maybe quite high vapor pressure, someone straight out of their degree program.  That turn over rate would be too high and it won’t work. So, you need a retention strategy.

 

Sunshine Hillygus – The one thing I worry about is when you point to an external review – is things get kicked down the road…   I worry about is there a way that we highlight the problem without just saying the solution is out there at some other institution.. and if we wait.. 

So, I want to flag that..

 

Tracy Futhey – I struggled with this…. In that I don’t want to present the answer, I want OIT to be the historian to the faculty process and feedback… I don’t see this report as prescriptive for fix problems we may have around our structures.

 

Sunshine Hillygus  – Highlight the identified the problem a little bit more.

 

Ken Rogerson – I’ve worked with both undergraduate honors thesis programs and graduates groups that use data. The resources can be the same but sometimes they are not. Maybe thinking a little bit more about differences between how we push or emphasize research between undergraduates and graduates.  

 

Tim McGeary – As a group that has data visualization that is a huge area that we cannot keep up with.

 

Terry Oas – Should this come from faculty, ITAC or OIT/ Tracy?

 

Tracy Futhey – Yes, I wrestled with this on where it should come from

 

Terry Oas -  It is possible that the powers that be could see that this as an OIT grab for resources…  so, this should come from faculty …

 

Mark Goodacre – It’s a phenomenal job… of doing this together.    It’s a pretty unique balancing act and I wonder if changing how it’s delivered can change the voice of who it’s coming from.

 

John Board and Jenny Lodge– It sounded like faculty wrote it.

 

Tracy Futhey – Appreciate your guidance; we will submit it as coming from ITAC / the faculty.