Incoming Faculty Orientation: Research Resources

Transcript

• 00:02Great. Thank you, and welcome
• 00:04to Yale, everybody.
• 00:06I'm here to present a
• 00:07few things that we have,
• 00:09developed recently as well as
• 00:11some continued resources
• 00:12for informatics,
• 00:14computing, data science, and
• 00:17AI at the school of
• 00:18medicine.
• 00:19So, my goals here
• 00:21are,
• 00:22or the goals of our
• 00:24core in as an informatics
• 00:25core to enable health discoveries
• 00:28through data analysis and computation,
• 00:29renovated here at the AIU.
• 00:31You will have, by the
• 00:32end of this calendar year,
• 00:33some state of the art,
• 00:36systems to compute, with your
• 00:38data
• 00:45and, do that in a
• 00:47secure privacy preserving HIPAA compliant
• 00:49manner.
• 00:50To support innovation with,
• 00:53by computational
• 00:54researchers, increase collaborations with all
• 00:56departments and the centers,
• 00:58in within and outside Yale,
• 01:00and then increased participation of,
• 01:02groups currently underrepresented
• 01:04in biomedical computing, which
• 01:07still, borrows a lot from
• 01:09engineering and that therefore has
• 01:10some of the same,
• 01:12issues of,
• 01:14unequal
• 01:15representation.
• 01:17And we do compute services
• 01:18and training. So I'll go
• 01:20briefly
• 01:21about our very colorful
• 01:23picture
• 01:24about where our compute environments
• 01:27reside. So the ones in
• 01:28the dotted lines,
• 01:30are within the health system,
• 01:32the Yale New Haven Health,
• 01:34and they are fully HIPAA
• 01:35compliant,
• 01:36and that's where all your
• 01:38human subjects
• 01:39research should be done at
• 01:41this point. It's the Yale,
• 01:43HIPAA compliant
• 01:45environment.
• 01:45There are two aspects of
• 01:47it. One is called,
• 01:48virtual desktop interface.
• 01:50That is when your,
• 01:52laptop serves as a terminal,
• 01:55so to speak, to a
• 01:57server that is with all
• 01:58the security
• 01:59out there. Much easier. You
• 02:01don't need to maintain your
• 02:02own applications. It's secure,
• 02:05and it's a way to,
• 02:07make sure that,
• 02:09if you inadvertently
• 02:11lose your laptop or so
• 02:12that you don't have to
• 02:13go through all the procedures
• 02:15that have to be done
• 02:17when you lose a a
• 02:18laptop that contains patient information.
• 02:21You probably know this from
• 02:23all your previous institutions,
• 02:25but I would like to
• 02:26emphasize
• 02:28in here
• 02:29that, you know, there is
• 02:30increased,
• 02:31attention to these matters.
• 02:34Also, there are some specs
• 02:36here very, crude, but, we
• 02:38have been adding GPUs to
• 02:40this environment so that you
• 02:42can do more,
• 02:44intense computation
• 02:46as well. And as you
• 02:47know,
• 02:48medical large language models, LLM,
• 02:52they require lots of GPUs,
• 02:53and we have some some
• 02:55versions of it planned for
• 02:57for new environments.
• 02:59Then we have something
• 03:00so ChipSafe
• 03:02is
• 03:02hardware from from here.
• 03:05Again,
• 03:06that,
• 03:07is directly at the servers
• 03:08that you can,
• 03:10compute, and you see on
• 03:11top several of the the
• 03:13logos for,
• 03:15programming language or environment
• 03:17or statistical
• 03:19packages that some of you
• 03:20may be familiar
• 03:21with.
• 03:23Now for,
• 03:24the number three, which is
• 03:25called spin up plus,
• 03:27that is Amazon Web Services,
• 03:30Amazon Cloud as many of
• 03:31you probably
• 03:32do your computation there.
• 03:35We have been,
• 03:36upgrading it to be a
• 03:37HIPAA compliant portion of the
• 03:39cloud for the Yale
• 03:41system. So, it's not there
• 03:43yet, but it will be
• 03:45pretty soon in which, you
• 03:47would compute as you compute
• 03:49with,
• 03:52your regular Amazon cloud,
• 03:54but with the peace of
• 03:56mind that the security and
• 03:58the privacy
• 03:59aspects are are being taken
• 04:00care for you.
• 04:02And the last one, again,
• 04:04the
• 04:05hardware will be installed, will
• 04:07be here all in December.
• 04:09We will do a bunch
• 04:10of testing and other things,
• 04:11and then the GPU nodes,
• 04:14NVIDIA ones will be available
• 04:16also,
• 04:17in a protected environment for,
• 04:20computation with LLMs, with AI,
• 04:23and all sorts of things,
• 04:25using patient data. So so
• 04:27I I do put here
• 04:29so that you know the
• 04:30environment,
• 04:31is being developed in pieces.
• 04:33Some of it is already
• 04:34there and some will be
• 04:35soon.
• 04:36For clinical research data services,
• 04:39so
• 04:40you're always,
• 04:41probably curious, how do I
• 04:42compute with data from the
• 04:44health system?
• 04:45So there are,
• 04:46free consultations
• 04:48on depending on what you
• 04:49need, training and office hours.
• 04:51There is a self-service
• 04:53data.
• 04:54So some of you who
• 04:55worked with Epic before may
• 04:57have worked with slicer dicer,
• 04:59which is an interface
• 05:00for population health, so to
• 05:03speak.
• 05:04And, you can do that
• 05:05or you can use freehand,
• 05:08SQL language
• 05:09that,
• 05:11it's often used for database
• 05:12retrieval, and you can, reuse
• 05:15existing reports and extract. So
• 05:17depending on your level
• 05:19of,
• 05:19expertise with data extractions and
• 05:22so on, you would utilize
• 05:23different types of services.
• 05:26There is also custom data
• 05:27engineering, data flow automations for
• 05:29projects that will require integration
• 05:32with epic services
• 05:34and few others things.
• 05:37Custom datasets and custom data
• 05:39marks. Some some, groups have
• 05:41registries,
• 05:42for example, of a particular,
• 05:44disease,
• 05:45all the patients in that
• 05:46and so on. So that
• 05:47all needs to be done
• 05:48in a very protected manner.
• 05:51Cosmos
• 05:52is,
• 05:53some of you may know
• 05:54or or or not, Epic
• 05:56has this consortium of institutions.
• 05:58There are two hundred forty
• 05:59seven,
• 06:01of them, I believe, across,
• 06:03the United States in which
• 06:05they agree to share
• 06:07the identified data or limited
• 06:09data sets, some some,
• 06:12special type of data. And
• 06:14only the members of the
• 06:15consortium can use. You are
• 06:17members as as you are
• 06:18at at Yale now, and
• 06:20then you can consult. And
• 06:21that's way more data than
• 06:23the local health system, which
• 06:25is a large one, but
• 06:27nothing compared to two hundred
• 06:29eighteen million patients.
• 06:31And also, as you know,
• 06:33the papers that are single
• 06:34center
• 06:36are less likely to be,
• 06:38published in high end journals
• 06:40and so on. So I
• 06:41encourage you
• 06:42to, get the training that
• 06:44is necessary for Cosmos. It's
• 06:46not everyone is allowed, but
• 06:48every department has,
• 06:51or will send a a
• 06:52representative to be the data
• 06:54person for Cosmos.
• 06:56So if you are interested,
• 06:57do ask your department representatives
• 07:00or come to us.
• 07:01OMOP is a standardized type
• 07:04of clinical data warehouse.
• 07:06It's analysis friendly friendly, and
• 07:08you can use,
• 07:09that as well.
• 07:12Clinical research applications, electronic data
• 07:15capture tools. You might have
• 07:16heard of REDCap. There is
• 07:18portions of EPIC for this
• 07:20purpose in in,
• 07:21another EDC.
• 07:24IRB approved econsent systems are
• 07:26available as well for your
• 07:28studies, clinical trial management, and
• 07:31integration with the electronic health
• 07:32record system.
• 07:35Integrated cohorts discovering screening
• 07:37management tools. You want to
• 07:39know how many patients with
• 07:40x, y, and z,
• 07:42have been seen in the
• 07:43system since
• 07:45two thousand and ten. That
• 07:47that should be something
• 07:48relatively easy to query.
• 07:51And then custom software engineering
• 07:53services that, takes longer. And
• 07:55then, again, it it's a
• 07:56discussion on is it doable,
• 07:58what are the timelines, what
• 08:00are the funding, and so
• 08:01on.
• 08:02Patient recruitment and outreach systems
• 08:05with a patient portal, self-service,
• 08:07cohort identification,
• 08:09or if it's complicated, let's
• 08:11say you have to have
• 08:13patients
• 08:14who have not taken this
• 08:16class of drugs
• 08:17before this procedure and and
• 08:20something that gets more complex,
• 08:22then you can also get
• 08:23assistance from data analysts.
• 08:26Like I said, create registries
• 08:28and self management,
• 08:30outreach lists.
• 08:31We also have the first,
• 08:34external database we have licensed
• 08:37for the whole Yale University
• 08:39is,
• 08:40Meritiv MarketScan.
• 08:42This is based,
• 08:44essentially of,
• 08:45claims data, and many people
• 08:47do research on that,
• 08:49as well, and it's available.
• 08:51We have partnered with the
• 08:53library so that they can
• 08:55help you with access to
• 08:56this, with access to the
• 08:58all of us,
• 09:00research program data, which,
• 09:02currently has a very large
• 09:04collection, the largest
• 09:05in the US
• 09:07of,
• 09:08sequences of human genomes. It
• 09:10has a lot of electronic
• 09:11health record data in standardized
• 09:13format,
• 09:14and it has,
• 09:16questionnaire data
• 09:17as well.
• 09:18And besides that, we have
• 09:20also,
• 09:21organized a local copy of
• 09:23the UK Biobank
• 09:25even though the the rules
• 09:26are changing for that Biobank
• 09:28and, will we
• 09:30be a similar model
• 09:32of the all of us
• 09:33in which you have to
• 09:34compute it in their particular
• 09:35designated
• 09:36cloud. But so far,
• 09:38we we have been able
• 09:40to,
• 09:41have a local version of
• 09:42it. And lastly, we have
• 09:45new online
• 09:46educational programs. Let me see
• 09:48if it shows here. In
• 09:50which what we did is
• 09:51recordings of, short modules. So
• 09:53if you were interested in
• 09:55a particular
• 09:56aspect of AI and medicine,
• 09:58we we developed a certificate
• 10:00program
• 10:01in which the the it's
• 10:02an online program in which
• 10:05every module is about a
• 10:07topic of interest. We are
• 10:09trying to set up a
• 10:10master's program in medical AI
• 10:13as well. And in many
• 10:15other trainees that I I
• 10:16don't have time here because,
• 10:19I gave sixty seconds to
• 10:21David Coleman this morning, so
• 10:23he has more time
• 10:25to talk to you today.
• 10:27So I'm not sure about
• 10:28the format, if you have
• 10:29any specific questions now or
• 10:31later.
• 10:32I maybe one question or
• 10:34two now.
• 10:37Or
• 10:38it was
• 10:39absolutely clear what what we
• 10:41have. Thank you so much.
• 10:43Thank you.
• 10:51And now we'll have doctor
• 10:52David Coleman, who's the director
• 10:53of YCCI.
• 10:55Let me just get back
• 10:57to yours. There we go.
• 10:59Thanks, Sam. Good morning, everyone.
• 11:02And,
• 11:03I'll join with the many
• 11:04other speakers in welcoming you
• 11:06and
• 11:07and,
• 11:08expressing how excited we are
• 11:10to have the
• 11:11opportunity,
• 11:12for you to join our
• 11:14community and to work with
• 11:15you.
• 11:17My role,
• 11:19here at Yale is to
• 11:20serve as the,
• 11:22current,
• 11:23interim director of the Yale
• 11:24Center for Clinical Investigation, as
• 11:26well as,
• 11:27director of clinical research for,
• 11:30the School of Medicine and
• 11:31Yale New Haven Health System,
• 11:33very recently,
• 11:35implemented,
• 11:36position.
• 11:38So I will give you
• 11:39a very, brief overview of
• 11:41some of the clinical research
• 11:42infrastructure
• 11:43at Yale in the hopes
• 11:45that,
• 11:46you will take advantage of
• 11:48what is a very, very
• 11:49substantial commitment of institutional
• 11:52resources to facilitate,
• 11:54discovery in in patients and
• 11:57humans.
• 11:59So,
• 12:02what I hope to cover
• 12:03is the, brief description of
• 12:05the infrastructure broadly, and then
• 12:08tell you in a little
• 12:09more detail about the Yale
• 12:10Center for Clinical Investigation,
• 12:13and then,
• 12:14describe some of the training
• 12:15options,
• 12:17available,
• 12:18to you as new faculty.
• 12:22So the clinical research,
• 12:24infrastructure
• 12:25at, the school of medicine
• 12:27is really
• 12:28divisible perhaps into three major
• 12:30groups. The first, the Yale
• 12:31Center for Clinical Investigation
• 12:34is the administrative,
• 12:36and intellectual home of the
• 12:38Clinical Translational Science Award,
• 12:40which,
• 12:41has been continuously
• 12:43awarded to,
• 12:45Yale,
• 12:46since the inception of the
• 12:48awards back in, two thousand
• 12:50six.
• 12:52We are a,
• 12:54a university center, that is
• 12:56we serve,
• 12:57all the schools of Yale
• 12:59University,
• 13:00although primarily,
• 13:02our our customers, if you
• 13:03will, are faculty and trainees
• 13:06from the School of Medicine.
• 13:09The Yale Cancer Center is
• 13:11a sister,
• 13:12organization
• 13:13that,
• 13:15is obviously focused on infrastructure
• 13:18and training for support
• 13:20of, cancer related studies. So
• 13:22it's reasonable to think of
• 13:24YCCI
• 13:25as primarily supporting,
• 13:27research and training
• 13:29in non oncology
• 13:31areas
• 13:32and the cancer center,
• 13:33obviously, focused on oncology.
• 13:36And then there have,
• 13:38over the,
• 13:40past fifteen or twenty years,
• 13:42been a number of departmentally
• 13:44based,
• 13:45in some cases, section based,
• 13:48units that provide,
• 13:50elements of clinical research,
• 13:53infrastructure.
• 13:54And I've listed here on
• 13:55this slide some of the
• 13:56departments that had made that,
• 13:59commitment.
• 14:01And so one has a,
• 14:03a variety of different,
• 14:05opportunities to be supported
• 14:07in your, research endeavors and
• 14:10your training,
• 14:12depending on the type of
• 14:13research you do, that is
• 14:14oncology versus non oncology,
• 14:17and the departments that you
• 14:18reside in.
• 14:20Our our center and the
• 14:21cancer centers fully support all
• 14:23of the departments that are
• 14:25on the the the third,
• 14:28bullet here.
• 14:29So it's not it's not
• 14:30that, if you're in cardiology,
• 14:32you can't get help from,
• 14:34either the cancer center or
• 14:36YCCI.
• 14:40So the mission of YCCI,
• 14:43is shown here. I'll go
• 14:45through it very quickly.
• 14:47This center has about two
• 14:48hundred and twenty five
• 14:50staff members.
• 14:52It was established in two
• 14:53thousand five,
• 14:55a precursor to the CTSA,
• 14:59RFA that that came out
• 15:00around that time.
• 15:03So our goals are really
• 15:06to accelerate
• 15:08and and really develop
• 15:10a a culture
• 15:11of,
• 15:12innovation,
• 15:13evaluation,
• 15:14collaboration,
• 15:15and implementation of new treatments,
• 15:19for, human disease.
• 15:22We work very carefully and,
• 15:24I think
• 15:25comprehensively
• 15:26with a number of community
• 15:27partners to co develop
• 15:29and implement,
• 15:31research that is particularly relevant,
• 15:34to improving
• 15:35the health outcomes of the
• 15:37communities served,
• 15:39by Yale New Haven Health
• 15:40System.
• 15:42We are,
• 15:44focusing very heavily on developing
• 15:46new
• 15:47translational research methods under the
• 15:49banner of,
• 15:51what, NIH is describing as
• 15:54translational
• 15:54science. So we are,
• 15:57very focused on supporting,
• 15:59new research methods to develop
• 16:02novel
• 16:03approaches to, clinical investigation.
• 16:07We have a very important
• 16:08training function.
• 16:10Our training programs, which I'll
• 16:12describe briefly,
• 16:14momentarily,
• 16:15are perhaps the most successful,
• 16:18demonstrably successful,
• 16:20elements of, YCCI's,
• 16:23now nearly twenty year history.
• 16:27And then,
• 16:28very recently, we have,
• 16:32been part of the,
• 16:34really invigorated
• 16:36commitment of the school of
• 16:38medicine
• 16:39and Yale New Haven Health
• 16:40System's leadership
• 16:42to really work in,
• 16:44in alignment in ways that
• 16:47that really ultimately benefit the
• 16:49patients that we serve and
• 16:51really promote,
• 16:52a level of,
• 16:54discovery that,
• 16:55extends
• 16:56geographically
• 16:57throughout,
• 16:58Connecticut and all of the
• 16:59areas served by the El
• 17:01Mohave and health system.
• 17:06We have many services that
• 17:08we offer,
• 17:09and I encourage you I'll
• 17:11I'll give you the website
• 17:13URL at the end. But,
• 17:15just to briefly,
• 17:17describe,
• 17:19we we,
• 17:20typically divide our services up
• 17:22into
• 17:23a pre award,
• 17:25or activation phase and then
• 17:28a,
• 17:29post award and study maintenance
• 17:30phase.
• 17:31In the pre award phase,
• 17:33we offer a number of
• 17:34services from,
• 17:36I have an idea,
• 17:39that I need to,
• 17:42refine in the form of
• 17:43a hypothesis
• 17:45and
• 17:45then design a research study
• 17:48to address that hypothesis.
• 17:50We have a number of
• 17:51people with expertise in study
• 17:53design
• 17:54and, the quantitative aspects of
• 17:57of constructing
• 17:58a meaningful clinical research protocol.
• 18:02And then we provide a
• 18:04lot of services related to
• 18:05getting a study,
• 18:08developed
• 18:09and activated,
• 18:11budgeted,
• 18:12and, off the ground.
• 18:14And then following the the
• 18:16the award of a research
• 18:19protocol,
• 18:20we work pretty,
• 18:22carefully,
• 18:24to provide a comprehensive range
• 18:27of services to help execute,
• 18:29the protocol.
• 18:31These range from,
• 18:33individuals to help with the
• 18:35various regulatory
• 18:36components to
• 18:38actual research,
• 18:39assistants,
• 18:41research nurses, research personnel,
• 18:43to,
• 18:44execute the,
• 18:47protocol.
• 18:53Our website
• 18:54is,
• 18:55undergoing very, very substantial revision.
• 18:58I just spent most of
• 18:59the morning,
• 19:00going through, the most recent
• 19:02draft of our website.
• 19:04We are migrating to some
• 19:06new software
• 19:07by mid September.
• 19:09We're going to be using
• 19:10Salesforce to provide a much
• 19:12higher level of timely service
• 19:15to the faculty.
• 19:16But,
• 19:17I just wanted to show
• 19:18you the the snapshot of
• 19:20our our current draft, which
• 19:22will have a,
• 19:24request services
• 19:25now,
• 19:26button,
• 19:27to link to the whole
• 19:29array of services that,
• 19:32YCCI,
• 19:33offers.
• 19:34So one can,
• 19:36really go on and and,
• 19:39select services,
• 19:41on an ad hoc basis
• 19:42or more comprehensive
• 19:44services related,
• 19:46to your,
• 19:47particular needs.
• 19:50And these forms are are
• 19:52all being,
• 19:53modified. So,
• 19:55the key is that, we
• 19:57wanna make it as easy
• 19:58as possible for you to
• 20:00access us, and we want,
• 20:02faculty to be able to
• 20:03come through a single portal,
• 20:05and we work then around
• 20:08that faculty member to provide,
• 20:10the services that are needed.
• 20:14So our, training resources
• 20:16are
• 20:17extremely important, and I urge
• 20:19you to take advantage of
• 20:21them. They range from,
• 20:23salaried positions on a in
• 20:25a, KL two format,
• 20:28to,
• 20:29various,
• 20:31seminars and courses,
• 20:33that can be very helpful
• 20:35to you as you,
• 20:38seek additional training.
• 20:40So in addition to the
• 20:41formal
• 20:42KL two program, which is
• 20:43a one or two year,
• 20:45award, it's,
• 20:47an institutional award through the
• 20:49CTSA.
• 20:51We offer, services and training
• 20:54in grant writing. This is
• 20:55one of our most popular
• 20:57and frankly successful offerings.
• 21:00We have a a program,
• 21:02so called ROCCATs program, which,
• 21:05is a, summer program in,
• 21:07developing collaborative skills and team
• 21:10science,
• 21:11a number of research and
• 21:13progress
• 21:14opportunities
• 21:14and and meetings,
• 21:16with other,
• 21:18trainees and scholars in the
• 21:19program.
• 21:21One of the most
• 21:22demonstrably
• 21:24consistent,
• 21:26sources of value of these
• 21:28training programs, and I I
• 21:29would emphasize this for you
• 21:31as people relatively new to
• 21:34this Yale community or at
• 21:35least, many of you are,
• 21:37Is is this these programs
• 21:39really create,
• 21:42networking
• 21:43and community.
• 21:44And
• 21:45we get,
• 21:47feedback consistently
• 21:49that
• 21:49from the participants
• 21:51say saying we learned a
• 21:52lot of content, but the
• 21:54best part of this was
• 21:55meeting other people
• 21:57who have now become friends
• 21:58and collaborators. So,
• 22:00as you think about your
• 22:02own strategy for integrating and
• 22:04taking advantage of what is
• 22:06a a remarkably
• 22:07talented university
• 22:09community,
• 22:10these training opportunities can be,
• 22:13really,
• 22:14important components of that strategy.
• 22:18We do also provide,
• 22:20assistance
• 22:21for,
• 22:23supplements, particularly diversity supplements.
• 22:26We,
• 22:28the grad school primarily,
• 22:30is the home for the
• 22:31investigative medicine program, which is
• 22:33Yale's version of,
• 22:35a MD to PhD,
• 22:38postgraduate training. It's an extremely
• 22:40successful
• 22:42program that's been around now
• 22:43for,
• 22:45two decades.
• 22:47We,
• 22:48support that with, a number
• 22:49of our faculty and our
• 22:51and our courses,
• 22:52and then we work arm
• 22:53and arm with the office
• 22:55of physician scientists and scientist
• 22:57development,
• 22:59to,
• 23:00provide,
• 23:01training opportunities
• 23:02for,
• 23:04faculty.
• 23:08Some of the the courses
• 23:10that are
• 23:11are offered are shown here.
• 23:12I won't go through all
• 23:13of them, but, simply to,
• 23:16suggest that you query our
• 23:18website,
• 23:19to identify some of these
• 23:21training opportunities because they are,
• 23:24quite valuable and either free
• 23:26or relatively inexpensive.
• 23:32So I will close then
• 23:34by,
• 23:35suggesting
• 23:36that,
• 23:38the university,
• 23:40the school,
• 23:41and the health system have
• 23:42made
• 23:43very substantial,
• 23:44and I would say durable,
• 23:47commitments
• 23:48to supporting high quality,
• 23:51clinical research
• 23:52and, research training.
• 23:55And I I hope,
• 23:57you enjoy the benefits of
• 23:59of those,
• 24:01commitments.
• 24:03Clinical research methods are evolving
• 24:06and,
• 24:07diversifying,
• 24:08and we are,
• 24:10participating
• 24:12in the collective efforts of
• 24:14this community to find new
• 24:16ways to discover,
• 24:17that make our results more
• 24:20generalizable
• 24:21and more impactful
• 24:22in positively,
• 24:24affecting health outcomes.
• 24:28I
• 24:29I wanna join with others
• 24:31in encouraging you,
• 24:33to take advantage of these
• 24:34research and training opportunities,
• 24:37and,
• 24:39to,
• 24:40reach out to me directly
• 24:42if I can help you
• 24:43and,
• 24:44others who are very, very
• 24:46invested in your,
• 24:48success.
• 24:49And then lastly, just to
• 24:51mention that,
• 24:52one of our challenges historically
• 24:54has been
• 24:55that,
• 24:56particularly new faculty, but faculty
• 24:58in general, have often not
• 25:00been as aware of the
• 25:02range of services
• 25:03that are offered.
• 25:05That is on us, and
• 25:07we are working diligently
• 25:08to improve,
• 25:09our outreach and our, web
• 25:12services and, and frankly, our
• 25:13usability,
• 25:15to to,
• 25:16be as helpful as we
• 25:18can be.
• 25:19So I will stop there
• 25:21and,
• 25:23ask for any questions or,
• 25:26comments that anyone has, any
• 25:27ways that
• 25:28I can clarify any of
• 25:30that content.