ASHG 2020 Abstract

[jimhavrilla]

Guidelines:

You are allotted 2,300 characters (excluding spaces) for the body text of your abstract. Title, author, and institutional data are not included in the 2,300 characters. See Step-by-Step Instructions (step 10) for details.

The docx (since I know that is the preference of the lab):

For now it reads like a pretty short tool paper. I'm not sure what else can be said about it in its present state and with the way queries work you can't just Google a phenotype and get our site first, we'd need to have pre-cached results for everything like GeneCards and MalaCards which I imagine costs them a fortune in server space.

PhenCards.docx

[kaichop]

The URL needs to be included, and an explicit description of input and output needs to be described. Ideally also a description what current other tools (such as HPO web server) cannot address

[jimhavrilla]

The thing is, MalaCards already does basically everything our site does. The main advantage we would provide over them is an API.

[jimhavrilla]

@kaichop I took a pass at expanding the output and input descriptions and added the URL. I also added a line about how we hope to improve on HPO. This is still an early draft so if there's more to add let me know. It's quite a bit longer now and reads as a deeper resource. Thank you for the notes. I'm still afraid to oversell it though given that MalaCards does all the same things. PhenCards_ASHG2020_Abstract.docx

[kaichop]

This is not the right attitude. You need to think about how to make yourself better, not that others are better and cannot be surpassed. Why spend the time and effort on something that you do not believe in? API is not an advantage and in reality few use it. It needs to be the go-to shop for phenotype information, more accurate, more comprehensive, and directly facilitates genetic analysis.

On Mon, Jun 8, 2020 at 6:55 PM Jim Havrilla notifications@github.com wrote:

The thing is, MalaCards https://www.malacards.org/card/burkitt_lymphoma already does basically everything our site does. The main advantage we would provide over them is an API.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640929408, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABNG3OFJIKH4BQGH7AX7NA3RVVT7XANCNFSM4NY2UGBQ .

[kaichop]

The abstract is written as if it is an addendum to HPO, which is clearly not the purpose or intention. Our goal is not to expand HPO resource. It is to create a go-to site for phenotype information and to facilitate genetic analysis.

On Mon, Jun 8, 2020 at 7:35 PM Jim Havrilla notifications@github.com wrote:

@kaichop https://github.com/kaichop I took a pass at expanding the output and input descriptions and added the URL. I also added a line about how we hope to improve on HPO. This is still an early draft so if there's more to add let me know. It's quite a bit longer now and reads as a deeper resource. Thank you for the notes. I'm still afraid to oversell it though given that MalaCards does all the same things. PhenCards_ASHG2020_Abstract.docx https://github.com/WGLab/Project_PhenCards/files/4749099/PhenCards_ASHG2020_Abstract.docx

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640941861, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABNG3OCKFJ4KPOYF2XQCH4LRVVYTNANCNFSM4NY2UGBQ .

[jimhavrilla]

Ok, I understand what you mean but in the past I was told not to oversell my work on abstracts because we have to deliver on it. I'm not sure if I can add, by myself, 70+ resources (that are already provided by Malacards) to compete with them by October. You're right that most will probably not use our API, but I think it could be useful for command line investigation. I'm not quite sure what we can do to compete and be that one-stop site for phenotypes. Sincerely, I am trying to think of ideas. What do you think? I am not trying to be difficult, lazy or negative, I'm just trying to think of what more we can do that has not been done without totally reworking our goals, maybe adding some way of combining this information comprehensively in a model (though they also provide a term graph structure). I think if we don't mention MalaCards the resource looks good but if we're trying to be competitive I'm not sure what we can offer. I agree that it's not the right attitude to assume they are going to be better, I'm not sure what to say in our abstract that can make ours sound better. Maybe I just don't understand your vision more concretely than a one-stop shop for phenotype information. I can't comprehend what that entails more than what has already been done. I think we can try to add all those same things and an API and Phen2Gene scores and make sure it is up-to-date but there are limited resources to add. Should we be creating predictive models? What is one example of something we can do to surpass such a comprehensive resource?

[kaichop]

The goal has always been one-stop shop for phenotype information, similar to what GeneCards is for genes. This has never been changed. Your stated goal in Abstract is not correct. It is not about over-selling phencards, but about correctly stating the goal. There are many innovations that can be considered, starting from existing ones such as integration with phen2gene and phenolyzer, to developing ones such as NLP-based gene/drug prediction from phenotype information (what Li has been working on), to novel psychiatric terms for ASD (or other development disorders) that Mengge is working on, to display of Phenlyzer network for every given phenotype term, to gene expression perturbation signatures for a phenotype (what Nasser has worked on before from LINCS), to human proteome project data on tissues of relevant to phenotypes, to model organism phenotype data and mouse knockout project phenotype data, to all GWAS catalogs that are relevant to the disease terms, to top ukbioank hits in various populations (rare or common variants) for a phenotype or icd code, ......, there are many creative ways how it can be more useful than malacards. I do not see what is the challenge for adding 70 or 100 or 200 databases, once an automation script is generated. Even Phenolyzer itself already incorporates a few dozen databases. Having more databases is important, but more importantly, having something useful that facilitates genetic analysis of human phenotype is what phencards should do.

On Mon, Jun 8, 2020 at 9:46 PM Jim Havrilla notifications@github.com wrote:

Ok, I understand what you mean but in the past I was told not to oversell my work on abstracts because we have to deliver on it. I'm not sure if I can add, by myself, 70+ resources (that are already provided by Malacards) to compete with them by October. You're right that most will probably not use our API, but I think it could be useful for command line investigation. I'm not quite sure what we can do to compete and be that one-stop site for phenotypes. Sincerely, I am trying to think of ideas. What do you think? I am not trying to be difficult, lazy or negative, I'm just trying to think of what more we can do that has not been done without totally reworking our goals, maybe adding some way of combining this information comprehensively in a model (though they also provide a term graph structure). I think if we don't mention MalaCards the resource looks good but if we're trying to be competitive I'm not sure what we can offer. I agree that it's not the right attitude to assume they are going to be better, I'm not sure what to say in our abstract that can make ours sound better. Maybe I just don't understand your vision more concretely than a one-stop shop for phenotype information. I can't comprehend what that entails more than what has already been done. I think we can try to add all those same things and an API and Phen2Gene scores and make sure it is up-to-date but there are limited resources to add. Should we be creating predictive models? What is one example of something we can do to surpass such a comprehensive resource?

On Mon, Jun 8, 2020, 9:17 PM Kai Wang notifications@github.com wrote:

The abstract is written as if it is an addendum to HPO, which is clearly not the purpose or intention. Our goal is not to expand HPO resource. It is to create a go-to site for phenotype information and to facilitate genetic analysis.

On Mon, Jun 8, 2020 at 7:35 PM Jim Havrilla notifications@github.com wrote:

@kaichop https://github.com/kaichop I took a pass at expanding the output and input descriptions and added the URL. I also added a line about how we hope to improve on HPO. This is still an early draft so if there's more to add let me know. It's quite a bit longer now and reads as a deeper resource. Thank you for the notes. I'm still afraid to oversell it though given that MalaCards does all the same things. PhenCards_ASHG2020_Abstract.docx <

https://github.com/WGLab/Project_PhenCards/files/4749099/PhenCards_ASHG2020_Abstract.docx

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <

https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640941861

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ABNG3OCKFJ4KPOYF2XQCH4LRVVYTNANCNFSM4NY2UGBQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640970712 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABSDYBFMDS35I25XR7ANKXLRVWERTANCNFSM4NY2UGBQ

.

Jim Havrilla

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640978620, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABNG3OCJETWNPFNMBVGR6ZTRVWH5ZANCNFSM4NY2UGBQ .

[jimhavrilla]

Okay, so maybe I should be focused on integrating our WGLab tools, i,e., EHR-Phenolyzer (or Doc2HPO), Phen2Gene, the novel psychiatric terms, gene expression information for a phenotype. Also should speak to those lab members all about integrating this data. And that's a good point about getting the PheWAS data in there. Once I have the UK Biobank results, that's another thing we can add. Would it be a good idea to try to link to Ben Neale's results too...? Hmm. GWAS Catalog is also a great idea. Thanks, I'll incorporate this into the writing of the abstract. I appreciate you giving me a more concrete idea of what you had in mind. Maybe it's an autism thing, but sometimes I can't see the big picture until I have an idea of the details.

I'm not sure what you mean by automation script, each resource has different formatting and sometimes requires a data transfer agreement to even obtain. So far we've had to write custom code for each resource and some have been impossible to download without manually making an account, moving to PC, then manually moving to cluster. There are numerous roadblocks to integrating so many resources.

On Mon, Jun 8, 2020 at 10:43 PM Kai Wang notifications@github.com wrote:

The goal has always been one-stop shop for phenotype information, similar to what GeneCards is for genes. This has never been changed. Your stated goal in Abstract is not correct. It is not about over-selling phencards, but about correctly stating the goal. There are many innovations that can be considered, starting from existing ones such as integration with phen2gene and phenolyzer, to developing ones such as NLP-based gene/drug prediction from phenotype information (what Li has been working on), to novel psychiatric terms for ASD (or other development disorders) that Mengge is working on, to display of Phenlyzer network for every given phenotype term, to gene expression perturbation signatures for a phenotype (what Nasser has worked on before from LINCS), to human proteome project data on tissues of relevant to phenotypes, to model organism phenotype data and mouse knockout project phenotype data, to all GWAS catalogs that are relevant to the disease terms, to top ukbioank hits in various populations (rare or common variants) for a phenotype or icd code, ......, there are many creative ways how it can be more useful than malacards. I do not see what is the challenge for adding 70 or 100 or 200 databases, once an automation script is generated. Even Phenolyzer itself already incorporates a few dozen databases. Having more databases is important, but more importantly, having something useful that facilitates genetic analysis of human phenotype is what phencards should do.

On Mon, Jun 8, 2020 at 9:46 PM Jim Havrilla notifications@github.com wrote:

Ok, I understand what you mean but in the past I was told not to oversell my work on abstracts because we have to deliver on it. I'm not sure if I can add, by myself, 70+ resources (that are already provided by Malacards) to compete with them by October. You're right that most will probably not use our API, but I think it could be useful for command line investigation. I'm not quite sure what we can do to compete and be that one-stop site for phenotypes. Sincerely, I am trying to think of ideas. What do you think? I am not trying to be difficult, lazy or negative, I'm just trying to think of what more we can do that has not been done without totally reworking our goals, maybe adding some way of combining this information comprehensively in a model (though they also provide a term graph structure). I think if we don't mention MalaCards the resource looks good but if we're trying to be competitive I'm not sure what we can offer. I agree that it's not the right attitude to assume they are going to be better, I'm not sure what to say in our abstract that can make ours sound better. Maybe I just don't understand your vision more concretely than a one-stop shop for phenotype information. I can't comprehend what that entails more than what has already been done. I think we can try to add all those same things and an API and Phen2Gene scores and make sure it is up-to-date but there are limited resources to add. Should we be creating predictive models? What is one example of something we can do to surpass such a comprehensive resource?

On Mon, Jun 8, 2020, 9:17 PM Kai Wang notifications@github.com wrote:

The abstract is written as if it is an addendum to HPO, which is clearly not the purpose or intention. Our goal is not to expand HPO resource. It is to create a go-to site for phenotype information and to facilitate genetic analysis.

On Mon, Jun 8, 2020 at 7:35 PM Jim Havrilla notifications@github.com wrote:

@kaichop https://github.com/kaichop I took a pass at expanding the output and input descriptions and added the URL. I also added a line about how we hope to improve on HPO. This is still an early draft so if there's more to add let me know. It's quite a bit longer now and reads as a deeper resource. Thank you for the notes. I'm still afraid to oversell it though given that MalaCards does all the same things. PhenCards_ASHG2020_Abstract.docx <

https://github.com/WGLab/Project_PhenCards/files/4749099/PhenCards_ASHG2020_Abstract.docx

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <

https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640941861

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ABNG3OCKFJ4KPOYF2XQCH4LRVVYTNANCNFSM4NY2UGBQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <

https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640970712

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ABSDYBFMDS35I25XR7ANKXLRVWERTANCNFSM4NY2UGBQ

.

Jim Havrilla

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640978620 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABNG3OCJETWNPFNMBVGR6ZTRVWH5ZANCNFSM4NY2UGBQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-640994775, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABSDYBBRGY7UXDLEV7FLBLDRVWOWJANCNFSM4NY2UGBQ .

[jimhavrilla]

OK I added some of your ideas. It's ok if you think it not good enough for ASHG 2020. It will definitely be a poster if we choose to submit. PhenCards_ASHG2020_Abstract.docx

[kaichop]

my change attached PhenCards_ASHG2020_Abstract KW.docx

[jimhavrilla]

These changes look good to me. Do you think it is good enough to submit after these changes? Or should I make one more pass?

On Wed, Jun 10, 2020 at 10:49 PM Kai Wang notifications@github.com wrote:

my change attached PhenCards_ASHG2020_Abstract KW.docx https://github.com/WGLab/Project_PhenCards/files/4762302/PhenCards_ASHG2020_Abstract.KW.docx

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-642373353, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABSDYBGD3L5AVD5WF23IFILRWBA4VANCNFSM4NY2UGBQ .

[kaichop]

good to submit, you have 48 hours after deadline to make changes, as I remember.

On Wed, Jun 10, 2020 at 11:02 PM Jim Havrilla notifications@github.com wrote:

These changes look good to me. Do you think it is good enough to submit after these changes? Or should I make one more pass?

On Wed, Jun 10, 2020 at 10:49 PM Kai Wang notifications@github.com wrote:

my change attached PhenCards_ASHG2020_Abstract KW.docx < https://github.com/WGLab/Project_PhenCards/files/4762302/PhenCards_ASHG2020_Abstract.KW.docx

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-642373353 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABSDYBGD3L5AVD5WF23IFILRWBA4VANCNFSM4NY2UGBQ

.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-642376694, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABNG3OB7S2KLBKGECAFQXKDRWBCMRANCNFSM4NY2UGBQ .

[jimhavrilla]

Ok thanks for the help.

On Wed, Jun 10, 2020 at 11:10 PM Kai Wang notifications@github.com wrote:

good to submit, you have 48 hours after deadline to make changes, as I remember.

On Wed, Jun 10, 2020 at 11:02 PM Jim Havrilla notifications@github.com wrote:

These changes look good to me. Do you think it is good enough to submit after these changes? Or should I make one more pass?

On Wed, Jun 10, 2020 at 10:49 PM Kai Wang notifications@github.com wrote:

my change attached PhenCards_ASHG2020_Abstract KW.docx <

https://github.com/WGLab/Project_PhenCards/files/4762302/PhenCards_ASHG2020_Abstract.KW.docx

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <

https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-642373353

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ABSDYBGD3L5AVD5WF23IFILRWBA4VANCNFSM4NY2UGBQ

.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-642376694 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABNG3OB7S2KLBKGECAFQXKDRWBCMRANCNFSM4NY2UGBQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/WGLab/Project_PhenCards/issues/35#issuecomment-642378727, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABSDYBG6PYATS35WJGLRU5TRWBDI5ANCNFSM4NY2UGBQ .

[jimhavrilla]

PhenCards_ASHG2020_Abstract KWclean.docx Submitted.