GO term clustering

[Ruiqi-CUB]

I got some very large GO enrichment figures even with strict cut-offs. There are over 1000 GO terms on the figure but a lot of them seem repetitive and way too specific. It's impossible to put those figures in my paper. Is it possible to "cluster" those GO terms to the next level or the next two levels based on the GO hierarchy?

[z0on]

Hey - this does not sound right. What do the fractions in front of the GO term names say? Can you paste a screenshot of part of the plotted tree (such that it would be readable)? Can you tell me more about your system: what is the test statistic you are basing your ranking on? (log-fold-change?) Misha

On Fri, Jan 27, 2023 at 1:07 PM Ruiqi-CUB @.***> wrote:

I got some very large GO enrichment figures even with strict cut-offs. There are over 1000 GO terms on the figure but a lot of them seem repetitive and way too specific. It's impossible to put those figures in my paper. Is it possible to "cluster" those GO terms to the next level or the next two levels based on the GO hierarchy?

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGEJ7D2YOEFBGJVUMFLWUQMGRANCNFSM6AAAAAAUJA247I . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[Ruiqi-CUB]

Thanks Misha. Here is a small portion of the figure on biological process: [image: image.png]

Organism: marine bivalve, LogFC=2, P=1e-3, the transcriptome was annotated with eggnog mapper. Here is a screenshot of the annotation result. As you can notice, there are lots of GO terms in each gene. [image: image.png]

On Fri, Jan 27, 2023 at 1:32 PM Mikhail V Matz @.***> wrote:

Hey - this does not sound right. What do the fractions in front of the GO term names say? Can you paste a screenshot of part of the plotted tree (such that it would be readable)? Can you tell me more about your system: what is the test statistic you are basing your ranking on? (log-fold-change?) Misha

On Fri, Jan 27, 2023 at 1:07 PM Ruiqi-CUB @.***> wrote:

I got some very large GO enrichment figures even with strict cut-offs. There are over 1000 GO terms on the figure but a lot of them seem repetitive and way too specific. It's impossible to put those figures in my paper. Is it possible to "cluster" those GO terms to the next level or the next two levels based on the GO hierarchy?

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABZUHGEJ7D2YOEFBGJVUMFLWUQMGRANCNFSM6AAAAAAUJA247I

. You are receiving this because you are subscribed to this thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1407045498, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALEILIWHXUAF6HM3IFGNKXTWUQWHBANCNFSM6AAAAAAUJA247I . You are receiving this because you authored the thread.Message ID: @.***>

-- Ruiqi Li PhD Candidate Dept. of Ecology and Evolutionary Biology University of Colorado Boulder [image: A button with "Hear my name" text for name playback in email signature] https://www.name-coach.com/ruiqi-li

[Ruiqi-CUB]

Thanks Misha. Here is a small portion of the figure on biological process:

Organism: marine bivalve, LogFC=2, P=1e-3, the transcriptome was annotated with eggnog mapper. Here is a screenshot of the annotation result. As you can notice, there are lots of GO terms in each gene.

[z0on]

Ha! This actually looks perfectly legit. The clearest immunity down-regulation I have ever seen :) To have fewer (only the most significant) terms displayed, I would simply adjust my plotting cutoffs, which are level1, level2, and level3 arguments for gomwuPlot: crank them down to 0.01, 0.003, 0.001 (play with these values until it looks good to you).

On Feb 6, 2023, at 9:44 AM, Ruiqi-CUB @.***> wrote:

Thanks Misha. Here is a small portion of the figure on biological process:

https://user-images.githubusercontent.com/46695842/217017079-d26d8f5a-1fa1-4cd5-83cc-415cfd205f9f.png Organism: marine bivalve, LogFC=2, P=1e-3, the transcriptome was annotated with eggnog mapper. Here is a screenshot of the annotation result. As you can notice, there are lots of GO terms in each gene. https://user-images.githubusercontent.com/46695842/217017120-0815ae71-bcbe-4890-9fce-d717a7e43386.png — Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1419296128, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGEF4WA7YDUEJKQMIY3WWEL7JANCNFSM6AAAAAAUJA247I. You are receiving this because you commented.

[Ruiqi-CUB]

Thanks a lot! Even after using strict cutoffs, the figure is still too long to put in any paper. Should I use the "clusterCutHeight=0.25" option in the gomwuStats function? Maybe use a really high value like 0.9?

On Mon, Feb 6, 2023 at 10:30 AM Mikhail V Matz @.***> wrote:

Ha! This actually looks perfectly legit. The clearest immunity down-regulation I have ever seen :) To have fewer (only the most significant) terms displayed, I would simply adjust my plotting cutoffs, which are level1, level2, and level3 arguments for gomwuPlot: crank them down to 0.01, 0.003, 0.001 (play with these values until it looks good to you).

On Feb 6, 2023, at 9:44 AM, Ruiqi-CUB @.***> wrote:

Thanks Misha. Here is a small portion of the figure on biological process:

< https://user-images.githubusercontent.com/46695842/217017079-d26d8f5a-1fa1-4cd5-83cc-415cfd205f9f.png

Organism: marine bivalve, LogFC=2, P=1e-3, the transcriptome was annotated with eggnog mapper. Here is a screenshot of the annotation result. As you can notice, there are lots of GO terms in each gene. < https://user-images.githubusercontent.com/46695842/217017120-0815ae71-bcbe-4890-9fce-d717a7e43386.png

— Reply to this email directly, view it on GitHub < https://github.com/z0on/GO_MWU/issues/15#issuecomment-1419296128>, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABZUHGEF4WA7YDUEJKQMIY3WWEL7JANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1419461346, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALEILIS5VJGZ3IVP5VV2ZBTWWEYLLANCNFSM6AAAAAAUJA247I . You are receiving this because you authored the thread.Message ID: @.***>

-- Ruiqi Li PhD Candidate Dept. of Ecology and Evolutionary Biology University of Colorado Boulder [image: A button with "Hear my name" text for name playback in email signature] https://www.name-coach.com/ruiqi-li

[z0on]

Hmm, can you just keep cranking these level1-level2-level3 cutoffs tighter and tighter until manageable number of GO terms remain? (I would not touch GO clustering cutoff, this may lead to poorly interpretable result)

On Feb 6, 2023, at 12:02 PM, Ruiqi-CUB @.***> wrote:

Thanks a lot! Even after using strict cutoffs, the figure is still too long to put in any paper. Should I use the "clusterCutHeight=0.25" option in the gomwuStats function? Maybe use a really high value like 0.9?

On Mon, Feb 6, 2023 at 10:30 AM Mikhail V Matz @.***> wrote:

Ha! This actually looks perfectly legit. The clearest immunity down-regulation I have ever seen :) To have fewer (only the most significant) terms displayed, I would simply adjust my plotting cutoffs, which are level1, level2, and level3 arguments for gomwuPlot: crank them down to 0.01, 0.003, 0.001 (play with these values until it looks good to you).

On Feb 6, 2023, at 9:44 AM, Ruiqi-CUB @.***> wrote:

Thanks Misha. Here is a small portion of the figure on biological process:

< https://user-images.githubusercontent.com/46695842/217017079-d26d8f5a-1fa1-4cd5-83cc-415cfd205f9f.png

Organism: marine bivalve, LogFC=2, P=1e-3, the transcriptome was annotated with eggnog mapper. Here is a screenshot of the annotation result. As you can notice, there are lots of GO terms in each gene. < https://user-images.githubusercontent.com/46695842/217017120-0815ae71-bcbe-4890-9fce-d717a7e43386.png

— Reply to this email directly, view it on GitHub < https://github.com/z0on/GO_MWU/issues/15#issuecomment-1419296128>, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABZUHGEF4WA7YDUEJKQMIY3WWEL7JANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1419461346, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALEILIS5VJGZ3IVP5VV2ZBTWWEYLLANCNFSM6AAAAAAUJA247I . You are receiving this because you authored the thread.Message ID: @.***>

-- Ruiqi Li PhD Candidate Dept. of Ecology and Evolutionary Biology University of Colorado Boulder [image: A button with "Hear my name" text for name playback in email signature] https://www.name-coach.com/ruiqi-li — Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1419517140, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGCQ3RHSXVS2JOLEBPDWWE4CXANCNFSM6AAAAAAUJA247I. You are receiving this because you commented.

[Ruiqi-CUB]

I have to use real small pvalues to make it manageable as you can see on the files attached. Would you still suggest using really tight cut-offs as I did in the 1e-7 figure instead of increase GO clustering cutoff? Thanks a lot. BP_SU_AM_vs_CM_1e-4.pdf BP_SU_AM_vs_CM_1e-7.pdf

[Ruiqi-CUB]

or should I play with largerst and smallest parameters in gomwuStats?

[z0on]

I’m confused: The picture you sent still shows default gomwuPlot levels (0.1, 0.05, 0.01)

Just in case, I have to ask: Did you put all analyzed genes into analysis (proper way), or just significant ones (wrong way)?

On Mon, Feb 6, 2023 at 10:43 PM Ruiqi-CUB @.***> wrote:

I have to use real small pvalues to make it manageable as you can see on the files attached. Would you still suggest using really tight cut-offs as I did in the 1e-7 figure instead of increase GO clustering cutoff? Thanks a lot. BP_SU_AM_vs_CM_1e-4.pdf https://github.com/z0on/GO_MWU/files/10671584/BP_SU_AM_vs_CM_1e-4.pdf BP_SU_AM_vs_CM_1e-7.pdf https://github.com/z0on/GO_MWU/files/10671585/BP_SU_AM_vs_CM_1e-7.pdf

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1420193760, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGFRDGF3QKZ2S4ZEIJDWWHHFLANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

-- cheers Misha matzlab.weebly.com

[Ruiqi-CUB]

Did you use the following links? They correctly shows the pvalues on my end. Yes I used all the genes from the transcriptome, not just DEGs.

BP_SU_AM_vs_CM_1e-4.pdf https://github.com/z0on/GO_MWU/files/10671584/BP_SU_AM_vs_CM_1e-4.pdf BP_SU_AM_vs_CM_1e-7.pdf https://github.com/z0on/GO_MWU/files/10671585/BP_SU_AM_vs_CM_1e-7.pdf

On Mon, Feb 6, 2023 at 10:24 PM Mikhail V Matz @.***> wrote:

I’m confused: The picture you sent still shows default gomwuPlot levels (0.1, 0.05, 0.01)

Just in case, I have to ask: Did you put all analyzed genes into analysis (proper way), or just significant ones (wrong way)?

On Mon, Feb 6, 2023 at 10:43 PM Ruiqi-CUB @.***> wrote:

I have to use real small pvalues to make it manageable as you can see on the files attached. Would you still suggest using really tight cut-offs as I did in the 1e-7 figure instead of increase GO clustering cutoff? Thanks a lot. BP_SU_AM_vs_CM_1e-4.pdf https://github.com/z0on/GO_MWU/files/10671584/BP_SU_AM_vs_CM_1e-4.pdf BP_SU_AM_vs_CM_1e-7.pdf https://github.com/z0on/GO_MWU/files/10671585/BP_SU_AM_vs_CM_1e-7.pdf

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1420193760, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABZUHGFRDGF3QKZ2S4ZEIJDWWHHFLANCNFSM6AAAAAAUJA247I

. You are receiving this because you commented.Message ID: @.***>

-- cheers Misha matzlab.weebly.com

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1420219710, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALEILISEAWX43IP4HOGPN7TWWHMBRANCNFSM6AAAAAAUJA247I . You are receiving this because you authored the thread.Message ID: @.***>

-- Ruiqi Li PhD Candidate Dept. of Ecology and Evolutionary Biology University of Colorado Boulder [image: A button with "Hear my name" text for name playback in email signature] https://www.name-coach.com/ruiqi-li

[z0on]

Sorry I missed those links. Ha, impressive. Never saw anything like that. Looks legit tho. Ok I agree, it makes sense increasing cutTreeHeight to merge all these terms more, as you suggested - did you try that yet? I’m very curious to see what happens.

On Mon, Feb 6, 2023 at 11:28 PM Ruiqi-CUB @.***> wrote:

Did you use the following links? They correctly shows the pvalues on my end. Yes I used all the genes from the transcriptome, not just DEGs.

BP_SU_AM_vs_CM_1e-4.pdf https://github.com/z0on/GO_MWU/files/10671584/BP_SU_AM_vs_CM_1e-4.pdf BP_SU_AM_vs_CM_1e-7.pdf https://github.com/z0on/GO_MWU/files/10671585/BP_SU_AM_vs_CM_1e-7.pdf

On Mon, Feb 6, 2023 at 10:24 PM Mikhail V Matz @.***> wrote:

I’m confused: The picture you sent still shows default gomwuPlot levels (0.1, 0.05, 0.01)

Just in case, I have to ask: Did you put all analyzed genes into analysis (proper way), or just significant ones (wrong way)?

On Mon, Feb 6, 2023 at 10:43 PM Ruiqi-CUB @.***> wrote:

I have to use real small pvalues to make it manageable as you can see on the files attached. Would you still suggest using really tight cut-offs as I did in the 1e-7 figure instead of increase GO clustering cutoff? Thanks a lot. BP_SU_AM_vs_CM_1e-4.pdf <https://github.com/z0on/GO_MWU/files/10671584/BP_SU_AM_vs_CM_1e-4.pdf

BP_SU_AM_vs_CM_1e-7.pdf <https://github.com/z0on/GO_MWU/files/10671585/BP_SU_AM_vs_CM_1e-7.pdf

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1420193760, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/ABZUHGFRDGF3QKZ2S4ZEIJDWWHHFLANCNFSM6AAAAAAUJA247I

. You are receiving this because you commented.Message ID: @.***>

-- cheers Misha matzlab.weebly.com

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1420219710, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ALEILISEAWX43IP4HOGPN7TWWHMBRANCNFSM6AAAAAAUJA247I

. You are receiving this because you authored the thread.Message ID: @.***>

-- Ruiqi Li PhD Candidate Dept. of Ecology and Evolutionary Biology University of Colorado Boulder [image: A button with "Hear my name" text for name playback in email signature] https://www.name-coach.com/ruiqi-li

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1420222343, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGFACJSYAUUJCATRO43WWHMQNANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

-- cheers Misha matzlab.weebly.com

[Ruiqi-CUB]

For the MF, using 0.25/0.5/0.75 clusterHeight gives me 161/118/92 GO terms at 10% FDR respectively; For the BP, it gives me 794/460/298 GO terms at 10% FDR respectively.

Here are the BP figures with 1e-3/1e-4/1e-5 cut-offs. As you can see, with strict cut-offs and higher clusterHeight parameter, I can get the GO terms to manageable size. Even with the high cutClusterHeight parameter(0.75), I think the trend is still the same. I am thinking of using high cutClusterHeight parameter(0.75) to represent the "big picture", but use small cutClusterHeight parameter(0.25) to get the "bestGOs", then go back to genes that might related to the process. Would you mind giving me some suggestions?

My study system is photosymbiotic cockles. Red terms are up-regulated in the light, blue ones are up-regulated in the dark.

Figures: BP_SU_AM_vs_CM_1e-5_clusterHeight0.5.pdf BP_SU_AM_vs_CM_1e-5_clusterHeight0.25.pdf BP_SU_AM_vs_CM_1e-5_clusterHeight0.75.pdf

[z0on]

I agree, good plan. I really wish I had your problem (too much of a functional signal). Cellular Component might look fun too (and sparser). BTW what fraction of "good genes" (those passing absValue cutoff) you are capturing with each version of the figure? That number is printed out to the screen by gomwuPlot Misha

On Tue, Feb 7, 2023 at 10:13 AM Ruiqi-CUB @.***> wrote:

For the MF, using 0.25/0.5/0.75 clusterHeight gives me 161/118/92 GO terms at 10% FDR respectively; For the BP, it gives me 794/460/298 GO terms at 10% FDR respectively.

Here are the BP figures with 1e-3/1e-4/1e-5 cut-offs. As you can see, with strict cut-offs and higher clusterHeight parameter, I can get the GO terms to manageable size. Even with the high cutClusterHeight parameter(0.75), I think the trend is still the same. I am thinking of using high cutClusterHeight parameter(0.75) to represent the "big picture", but use small cutClusterHeight parameter(0.25) to get the "bestGOs", then go back to genes that might related to the process. Would you mind giving me some suggestions?

My study system is photosymbiotic cockles. Red terms are up-regulated in the light, blue ones are up-regulated in the dark.

Figures: BP_SU_AM_vs_CM_1e-5_clusterHeight0.5.pdf https://github.com/z0on/GO_MWU/files/10677031/BP_SU_AM_vs_CM_1e-5_clusterHeight0.5.pdf BP_SU_AM_vs_CM_1e-5_clusterHeight0.25.pdf https://github.com/z0on/GO_MWU/files/10677033/BP_SU_AM_vs_CM_1e-5_clusterHeight0.25.pdf BP_SU_AM_vs_CM_1e-5_clusterHeight0.75.pdf https://github.com/z0on/GO_MWU/files/10677149/BP_SU_AM_vs_CM_1e-5_clusterHeight0.75.pdf

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421035411, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGDBFAHWIWE7NCVLCOTWWJYCVANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

[Ruiqi-CUB]

clusterHeight0.25: 1044 out of 1793 ( 58% ) clusterHeight0.5: 1052 out of 1793 ( 59% ) clusterHeight0.75: 566 out of 1084 ( 52% )

Should I concern about the reduce of total number of good genes in clusterHeight0.75?

[z0on]

hmm! I am surprised that good genes' number changes, as it is supposed to be the property of your dataset and should not depend on the way you analyze it.... there might be a bug somewhere. Can you please count what is the actual number of "good genes" in your input data?

On Tue, Feb 7, 2023 at 11:33 AM Ruiqi-CUB @.***> wrote:

clusterHeight0.25: 1044 out of 1793 ( 58% ) clusterHeight0.5: 1052 out of 1793 ( 59% ) clusterHeight0.75: 566 out of 1084 ( 52% )

Should I concern about the reduce of total number of good genes in clusterHeight0.75?

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421168756, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGBTHIWMW2CHVYYBZEDWWKBNBANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

[Ruiqi-CUB]

Sorry I just found that I made a mistake: I forgot the unmark the "absValue=-log(0.05,10)" for clusterHeight0.75.

clusterHeight0.75: "Good genes" accounted for: 940 out of 1793 ( 52% ). Here is the new figure (only the number of good candidates changes): BP_SU_AM_vs_CM_1e-5_clusterHeight0.75.pdf

[Ruiqi-CUB]

There are 6229 genes whose |logFC|>1 in my input data. Here are the CC figures by the way. CC_SU_AM_vs_CM_1e-3_clusterHeight0.25.pdf CC_SU_AM_vs_CM_1e-3_clusterHeight0.5.pdf CC_SU_AM_vs_CM_1e-3_clusterHeight0.75.pdf I think they are still showing the same trends.

[z0on]

but the question is. how many of these 6229 are annotated with GO terms for this specific GO division?... I see it is not so easy to just count :)

On Tue, Feb 7, 2023 at 12:15 PM Ruiqi-CUB @.***> wrote:

There are 6229 genes whose |logFC|>1 in my input data. Here are the CC figures by the way. CC_SU_AM_vs_CM_1e-3_clusterHeight0.25.pdf https://github.com/z0on/GO_MWU/files/10678165/CC_SU_AM_vs_CM_1e-3_clusterHeight0.25.pdf CC_SU_AM_vs_CM_1e-3_clusterHeight0.5.pdf https://github.com/z0on/GO_MWU/files/10678170/CC_SU_AM_vs_CM_1e-3_clusterHeight0.5.pdf CC_SU_AM_vs_CM_1e-3_clusterHeight0.75.pdf https://github.com/z0on/GO_MWU/files/10678172/CC_SU_AM_vs_CM_1e-3_clusterHeight0.75.pdf I think they are still showing the same trends.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421239417, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGBPZZ73EGSHKATSUVLWWKGKNANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

[z0on]

I see you have chloroplast components here, so it is obviously not just the cockle genome response we are looking at... Did you map to a combined cockle+symbiont reference? If yes, how do you account for potentially varying symbiont density (which would look like changes in all symbiont genes)?

On Tue, Feb 7, 2023 at 12:15 PM Ruiqi-CUB @.***> wrote:

There are 6229 genes whose |logFC|>1 in my input data. Here are the CC figures by the way. CC_SU_AM_vs_CM_1e-3_clusterHeight0.25.pdf https://github.com/z0on/GO_MWU/files/10678165/CC_SU_AM_vs_CM_1e-3_clusterHeight0.25.pdf CC_SU_AM_vs_CM_1e-3_clusterHeight0.5.pdf https://github.com/z0on/GO_MWU/files/10678170/CC_SU_AM_vs_CM_1e-3_clusterHeight0.5.pdf CC_SU_AM_vs_CM_1e-3_clusterHeight0.75.pdf https://github.com/z0on/GO_MWU/files/10678172/CC_SU_AM_vs_CM_1e-3_clusterHeight0.75.pdf I think they are still showing the same trends.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421239417, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGBPZZ73EGSHKATSUVLWWKGKNANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

[Ruiqi-CUB]

I did mapped to a bivale and a Symbiodiniaceae database respectively, genes mapped to Symbiodiniaceae database but not bivalve database were removed. Techinically I am only looking at the host genes. I think terms like "chloroplast" might just be noise as those GO terms are pretty general. Lots of them don't make sense in my study system.

I am using grep to get the number of DE Genes that have GO annotation. It just takes a well.

[z0on]

Well, chloroplast genes are not really generic, animals are not supposed to have them - I would be worried.

On Tue, Feb 7, 2023 at 12:53 PM Ruiqi-CUB @.***> wrote:

I did mapped to a bivale and a Symbiodiniaceae database respectively, genes mapped to Symbiodiniaceae database but not bivalve database were removed. Techinically I am only looking at the host genes. I think terms like "chloroplast" might just be noise as those GO terms are pretty general. Lots of them don't make sense in my study system.

I am using grep to get the number of DE Genes that have GO annotation. It just takes a well.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421286778, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGHOWG7ZO3H5Q3NVUFDWWKK4BANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

[Ruiqi-CUB]

I will double check the filtering. Thanks a lot. I got 1926 genes have GOs and |logFC|>1, which is not consistent with 1793, but close to.

[z0on]

yes but it has to be the specific GO division that you are running through GO_MWU - CC, MF, or BP

On Tue, Feb 7, 2023 at 3:28 PM Ruiqi-CUB @.***> wrote:

I will double check the filtering. Thanks a lot. I got 1926 genes have GOs and |logFC|>1, which is not consistent with 1793, but close to.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421476881, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGEXM775L33TUT5EDXLWWK47FANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

[Ruiqi-CUB]

Is there a way to do the count for the specific category? I can't think of an easy way to do it.

On Tue, Feb 7, 2023 at 3:12 PM Mikhail V Matz @.***> wrote:

yes but it has to be the specific GO division that you are running through GO_MWU - CC, MF, or BP

On Tue, Feb 7, 2023 at 3:28 PM Ruiqi-CUB @.***> wrote:

I will double check the filtering. Thanks a lot. I got 1926 genes have GOs and |logFC|>1, which is not consistent with 1793, but close to.

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421476881, or unsubscribe < https://github.com/notifications/unsubscribe-auth/ABZUHGEXM775L33TUT5EDXLWWK47FANCNFSM6AAAAAAUJA247I

. You are receiving this because you commented.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1421532085, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALEILIVHLVTSGBAL6JNHPXLWWLCGTANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

-- Ruiqi Li PhD Candidate Dept. of Ecology and Evolutionary Biology University of Colorado Boulder [image: A button with "Hear my name" text for name playback in email signature] https://www.name-coach.com/ruiqi-li

[Ruiqi-CUB]

Hi Misha, Onces I have the "best GOs", is there an easier way to get the "good genes" using the GO terms? Thanks Ruiqi

[z0on]

Let me worry about the good genes count (that is pretty irrelevant in the grand scheme of things since it does not affect go-mwu stats, I just need to see if I miscount them somehow) , you worry about chloroplast genes!

On Wed, Feb 8, 2023 at 10:31 PM Ruiqi-CUB @.***> wrote:

Hi Misha, Onces I have the "best GOs", is there an easier way to get the "good genes" using the GO terms? Thanks Ruiqi

— Reply to this email directly, view it on GitHub https://github.com/z0on/GO_MWU/issues/15#issuecomment-1423618482, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABZUHGG5KWTKW4QLZIK4KQDWWRXIZANCNFSM6AAAAAAUJA247I . You are receiving this because you commented.Message ID: @.***>

-- cheers Misha matzlab.weebly.com