From segway-users at u.washington.edu Wed Aug 13 15:49:37 2025 From: segway-users at u.washington.edu (JOSEPH KIRANGWA via Segway-users) Date: Wed Aug 13 15:52:42 2025 Subject: [Segway-users] Is it possible to segment the genome with one model for two conditions? Message-ID: Hi, I am interested in genomic segments for two conditions learned by one model. I have tried generating combined genome data file and training the model on both conditions. Then use Segway identify with specific genomedata archive each condition with the above generated traindir but this gives an error out of range matrix. I need two separate bed files for each condition but generated using one learned model. This is then easy to study differences in states between the two conditions. How to achieve this in Segway? Thanks a lot. *Joseph * *Postdoctoral Research Associate* Department of Biochemistry | University of Oxford Dorothy Crowfoot Hodgkin Building | South Parks Rd | OX1 3QU *Phone: +447979529878* *Email: joseph.kirangwa@bioch.ox.ac.uk * *https://psarkies.wixsite.com/epievo * -------------- next part -------------- An HTML attachment was scrubbed... URL: From segway-users at u.washington.edu Thu Aug 14 06:08:29 2025 From: segway-users at u.washington.edu (Roberts, Eric via Segway-users) Date: Thu Aug 14 08:59:21 2025 Subject: [Segway-users] [External] Is it possible to segment the genome with one model for two conditions? In-Reply-To: References: Message-ID: Hello, Yes, two different input datasets should be enough to segment the genome if that's what you mean by conditions. How exactly are you running the genome? Are you creating a Genomedata archive from two BED files or are you using the BED files for something else? What does your input data look like? If you could be more specific about your command for running Segway and provide the exact error message output we can probably figure out how to address your issue. Eric ________________________________ From: Segway-users on behalf of JOSEPH KIRANGWA via Segway-users Sent: Wednesday, August 13, 2025 6:49 PM To: segway-users@uw.edu Subject: [External] [Segway-users] Is it possible to segment the genome with one model for two conditions? Hi, I am interested in genomic segments for two conditions learned by one model. I have tried generating combined genome data file and training the model on both conditions. Then use Segway identify with specific genomedata archive each condition with the above generated traindir but this gives an error out of range matrix. I need two separate bed files for each condition but generated using one learned model. This is then easy to study differences in states between the two conditions. How to achieve this in Segway? Thanks a lot. Joseph Postdoctoral Research Associate Department of Biochemistry | University of Oxford Dorothy Crowfoot Hodgkin Building | South Parks Rd | OX1 3QU Phone: +447979529878 Email: joseph.kirangwa@bioch.ox.ac.uk https://psarkies.wixsite.com/epievo This e-mail may contain confidential and/or privileged information for the sole use of the intended recipient. Any review or distribution by anyone other than the person for whom it was originally intended is strictly prohibited. If you have received this e-mail in error, please contact the sender and delete all copies. Opinions, conclusions or other information contained in this e-mail may not be that of the organization. If you feel you have received an email from UHN of a commercial nature and would like to be removed from the sender's mailing list please do one of the following: (1) Follow any unsubscribe process the sender has included in their email (2) Where no unsubscribe process has been included, reply to the sender and type "unsubscribe" in the subject line. If you require additional information please go to our UHN Newsletters and Mailing Lists page. Please note that we are unable to automatically unsubscribe individuals from all UHN mailing lists. Patient Consent for Email: UHN patients may provide their consent to communicate with UHN about their care using email. All electronic communication carries some risk. Please visit the UHN website and search "Using email to Communicate with Your UHN Health Care Team" to learn about the risks of electronic communication and how to protect your privacy. You may withdraw your consent to receive emails from UHN at any time. Please contact your care provider, if you do not wish to receive emails from UHN. -------------- next part -------------- An HTML attachment was scrubbed... URL: From segway-users at u.washington.edu Thu Aug 14 09:54:38 2025 From: segway-users at u.washington.edu (JOSEPH KIRANGWA via Segway-users) Date: Fri Aug 15 14:53:38 2025 Subject: [Segway-users] [External] Is it possible to segment the genome with one model for two conditions? In-Reply-To: References: Message-ID: Dear Eric, Thanks for your reply. Here are the three possible scenarios and I am interested in (b) ; (a) Multiple cell types are treated independently, leading to a different model and annotation for each cell type. (I am against this because we are interested in doing joint analysis of the data downstream) (b) *Multiple cell types (conditions) are effectively concatenated, leading to one shared model for all cell types (conditions), but cell-type (condition)-specific annotations. This is what I want to do with Segway.* (c) Data from multiple cell types are stacked, leading to one model based on an expanded set of features, and one annotation of the genome. (*This seems be the default in Segway*) Please find the attached analysis that I have tried. Please advise me on how to proceed. I am using the version, segway 3.0.4. Thanks a lot. Kind regards, *Joseph * *Postdoctoral Research Associate* Department of Biochemistry | University of Oxford Dorothy Crowfoot Hodgkin Building | South Parks Rd | OX1 3QU *Phone: +447979529878* On Thu, Aug 14, 2025 at 2:08?PM Roberts, Eric wrote: > Hello, > > Yes, two different input datasets should be enough to segment the genome > if that's what you mean by conditions. > > How exactly are you running the genome? Are you creating a Genomedata > archive from two BED files or are you using the BED files for something > else? What does your input data look like? > > If you could be more specific about your command for running Segway and > provide the exact error message output we can probably figure out how to > address your issue. > > Eric > ------------------------------ > *From:* Segway-users on > behalf of JOSEPH KIRANGWA via Segway-users > *Sent:* Wednesday, August 13, 2025 6:49 PM > *To:* segway-users@uw.edu > *Subject:* [External] [Segway-users] Is it possible to segment the genome > with one model for two conditions? > > Hi, > > I am interested in genomic segments for two conditions learned by one > model. > > I have tried generating combined genome data file and training the model > on both conditions. > > Then use Segway identify with specific genomedata archive each condition > with the above generated traindir but this gives an error out of range > matrix. > > I need two separate bed files for each condition but generated using one > learned model. > > This is then easy to study differences in states between the two > conditions. > > How to achieve this in Segway? > > Thanks a lot. > > *Joseph * > *Postdoctoral Research Associate* > > Department of Biochemistry | University of Oxford > > Dorothy Crowfoot Hodgkin Building | South Parks Rd | OX1 3QU > *Phone: +447979529878* > > *Email: joseph.kirangwa@bioch.ox.ac.uk * > *https://psarkies.wixsite.com/epievo > * > > > > > > > > This e-mail may contain confidential and/or privileged information for the > sole use of the intended recipient. > Any review or distribution by anyone other than the person for whom it was > originally intended is strictly prohibited. > If you have received this e-mail in error, please contact the sender and > delete all copies. > Opinions, conclusions or other information contained in this e-mail may > not be that of the organization. > > If you feel you have received an email from UHN of a commercial nature and > would like to be removed from the sender's mailing list please do one of > the following: > (1) Follow any unsubscribe process the sender has included in their email > (2) Where no unsubscribe process has been included, reply to the sender > and type "unsubscribe" in the subject line. If you require additional > information please go to our UHN Newsletters and Mailing Lists page. > Please note that we are unable to automatically unsubscribe individuals > from all UHN mailing lists. > > Patient Consent for Email: > > UHN patients may provide their consent to communicate with UHN about their > care using email. All electronic communication carries some risk. Please > visit the UHN website and search ?Using email to Communicate with Your UHN > Health Care Team? to learn about the risks of electronic communication and > how to protect your privacy. You may withdraw your consent to receive > emails from UHN at any time. Please contact your care provider, if you do > not wish to receive emails from UHN. > -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- #Generate genomedata for both Parasitic and Freeliving females genomedata-load \ -s s_ratti.SRAEv7.genome.v7.0.fa \ -t FreeLiving_H3K27me3=Free-Living-female_H3K27me3.bedGraph \ -t FreeLiving_H3K36me3=Free-Living-female_H3K36me3.bedGraph \ -t FreeLiving_H3K9me3=Free-Living-female_H3K9me3.bedGraph \ -t FreeLiving_Input=Free-Living-female_Input.bedGraph \ -t Parasitic_H3K27me3=Parasitic-female_H3K27me3.bedGraph \ -t Parasitic_H3K36me3=Parasitic-female_H3K36me3.bedGraph \ -t Parasitic_H3K9me3=Parasitic-female_H3K9me3.bedGraph \ -t Parasitic_Input=Parasitic-female_Input.bedGraph \ combined_conditions.genomedata #Segway train segway train --num-labels=6 combined_conditions.genomedata traindir segway identify combined_conditions.genomedata traindir identifydir #This works but generates segway.bed for both parasitic and non parasitic) #What we want is to be able to learn one model and segment the genome to generate two bed files, one for parasitic and another #for freeliving female #Parasitic female (condition 1) genomedata-load \ -s s_ratti.SRAEv7.genome.v7.0.fa \ -t Parasitic_H3K27me3=Parasitic-female_H3K27me3.bedGraph \ -t Parasitic_H3K9me3=Parasitic-female_H3K9me3.bedGraph \ -t Parasitic_H3K36me3=Parasitic-female_H3K36me3.bedGraph \ -t Parasitic_Input=Parasitic-female_Input.bedGraph \ Parasitic-female.genomedata #Annotate the parasitic female with (Interested in parasitic_female_segway.bed) #This does not work #traindir is from the (segway train --num-labels=6 combined_conditions.genomedata traindir) segway identify parasitic_female.genomedata traindir identify_parasitic_female 2> segway_error.log #Freeliving female (condition 2) genomedata-load \ -s s_ratti.SRAEv7.genome.v7.0.fa \ -t FreeLiving_H3K27me3=Free-Living-female_H3K27me3.bedGraph \ -t FreeLiving_H3K9me3=Free-Living-female_H3K9me3.bedGraph \ -t FreeLiving_H3K36me3=Free-Living-female_H3K36me3.bedGraph \ -t FreeLiving_Input=Free-Living-female_Input.bedGraph \ Free-Living-female.genomedata #identify/Annotate segway identify Free-Living-female.genomedata traindir identify_Free-Living-female #Notes #We are interested in learning one model for both parasitic and freeliving female to understand how chromatin states change #Basically to generate two bed files (parasitic_female_segway.bed and Free-Living-female_segway.bed). Using the same model to #learn both conditions and generate seperate segmentations for each condition. This can then be comparable for #downstream analysis using tools such as bedtools intersect. From segway-users at u.washington.edu Mon Aug 18 12:58:08 2025 From: segway-users at u.washington.edu (Roberts, Eric via Segway-users) Date: Mon Aug 18 13:31:30 2025 Subject: [Segway-users] [External] Is it possible to segment the genome with one model for two conditions? In-Reply-To: References: Message-ID: If you want to run in concatenated mode to train a single model across both datasets you can seperate the tracks by commas. See https://segway.readthedocs.io/en/latest/segway.html#tracks For this I would use the Genomedata archive you created in your first step mentioned in your attachment containing all the tracks. If the concatenated mode doesn't work for you or there are further errors let us know. Eric ________________________________ From: JOSEPH KIRANGWA Sent: Thursday, August 14, 2025 12:54 PM To: Roberts, Eric Cc: segway-users@uw.edu Subject: Re: [External] [Segway-users] Is it possible to segment the genome with one model for two conditions? Dear Eric, Thanks for your reply. Here are the three possible scenarios and I am interested in (b) ; (a) Multiple cell types are treated independently, leading to a different model and annotation for each cell type. (I am against this because we are interested in doing joint analysis of the data downstream) (b) Multiple cell types (conditions) are effectively concatenated, leading to one shared model for all cell types (conditions), but cell-type (condition)-specific annotations. This is what I want to do with Segway. (c) Data from multiple cell types are stacked, leading to one model based on an expanded set of features, and one annotation of the genome. (This seems be the default in Segway) Please find the attached analysis that I have tried. Please advise me on how to proceed. I am using the version, segway 3.0.4. Thanks a lot. Kind regards, Joseph Postdoctoral Research Associate Department of Biochemistry | University of Oxford Dorothy Crowfoot Hodgkin Building | South Parks Rd | OX1 3QU Phone: +447979529878 On Thu, Aug 14, 2025 at 2:08?PM Roberts, Eric > wrote: Hello, Yes, two different input datasets should be enough to segment the genome if that's what you mean by conditions. How exactly are you running the genome? Are you creating a Genomedata archive from two BED files or are you using the BED files for something else? What does your input data look like? If you could be more specific about your command for running Segway and provide the exact error message output we can probably figure out how to address your issue. Eric ________________________________ From: Segway-users > on behalf of JOSEPH KIRANGWA via Segway-users > Sent: Wednesday, August 13, 2025 6:49 PM To: segway-users@uw.edu > Subject: [External] [Segway-users] Is it possible to segment the genome with one model for two conditions? Hi, I am interested in genomic segments for two conditions learned by one model. I have tried generating combined genome data file and training the model on both conditions. Then use Segway identify with specific genomedata archive each condition with the above generated traindir but this gives an error out of range matrix. I need two separate bed files for each condition but generated using one learned model. This is then easy to study differences in states between the two conditions. How to achieve this in Segway? Thanks a lot. Joseph Postdoctoral Research Associate Department of Biochemistry | University of Oxford Dorothy Crowfoot Hodgkin Building | South Parks Rd | OX1 3QU Phone: +447979529878 Email: joseph.kirangwa@bioch.ox.ac.uk https://psarkies.wixsite.com/epievo This e-mail may contain confidential and/or privileged information for the sole use of the intended recipient. Any review or distribution by anyone other than the person for whom it was originally intended is strictly prohibited. If you have received this e-mail in error, please contact the sender and delete all copies. Opinions, conclusions or other information contained in this e-mail may not be that of the organization. If you feel you have received an email from UHN of a commercial nature and would like to be removed from the sender's mailing list please do one of the following: (1) Follow any unsubscribe process the sender has included in their email (2) Where no unsubscribe process has been included, reply to the sender and type "unsubscribe" in the subject line. If you require additional information please go to our UHN Newsletters and Mailing Lists page. Please note that we are unable to automatically unsubscribe individuals from all UHN mailing lists. Patient Consent for Email: UHN patients may provide their consent to communicate with UHN about their care using email. All electronic communication carries some risk. Please visit the UHN website and search ?Using email to Communicate with Your UHN Health Care Team? to learn about the risks of electronic communication and how to protect your privacy. You may withdraw your consent to receive emails from UHN at any time. Please contact your care provider, if you do not wish to receive emails from UHN. This e-mail may contain confidential and/or privileged information for the sole use of the intended recipient. Any review or distribution by anyone other than the person for whom it was originally intended is strictly prohibited. If you have received this e-mail in error, please contact the sender and delete all copies. Opinions, conclusions or other information contained in this e-mail may not be that of the organization. If you feel you have received an email from UHN of a commercial nature and would like to be removed from the sender's mailing list please do one of the following: (1) Follow any unsubscribe process the sender has included in their email (2) Where no unsubscribe process has been included, reply to the sender and type "unsubscribe" in the subject line. If you require additional information please go to our UHN Newsletters and Mailing Lists page. Please note that we are unable to automatically unsubscribe individuals from all UHN mailing lists. Patient Consent for Email: UHN patients may provide their consent to communicate with UHN about their care using email. All electronic communication carries some risk. Please visit the UHN website and search ?Using email to Communicate with Your UHN Health Care Team? to learn about the risks of electronic communication and how to protect your privacy. You may withdraw your consent to receive emails from UHN at any time. Please contact your care provider, if you do not wish to receive emails from UHN. -------------- next part -------------- An HTML attachment was scrubbed... URL: