Commonly methylated

[freedomq8]

Hi there, Thanks for keeping up with my questions. As bicycle is able to from differential analysis between two different groups i.e treatment vs control. I wanted to know if bicycle can perform commonly methylated CpG shores, CPG islands, CpG shelves and promoter regions among a group of samples i.e control samples ?

Many thanks

[osvaldogc]

Hello,

Bicycle is able to check methylation levels at CpG sites in all those regions defined by the user: At the moment, bicycle does not contain or provide annotation files with specific definitions of methylation sites (CpG shores and shelves, promoter regions, etc).

That is, in part, one of its main forces, as you can provide as many specific annotations sets as you wan to provide, and bicycle will investigate methylation levels in all those specified regions. For instance, you must provide the genome annotation of all gene promoter regions in the genome (in BED file format), in order for bicycle to investigate methylation levels within those specific regions. The genomic coordinates for promoter regions are easy to obtain from a genome annotation file (depending on your definition of promoter, for instance, 1500bps<TSS<500bps). This would be similar for CpG shores and shelves, CpG islands. If you provide different annotation files to bicycle (several bed files at once), it will measure methylation levels in all those files regions, and perform differential methylation tests for all of them.

I hope this helps.

osvaldo.

---

From: freedomq8 [notifications@github.com] Sent: Wednesday, April 22, 2020 11:06 AM To: sing-group/bicycle Cc: Subscribed Subject: [sing-group/bicycle] Commonly methylated (#5)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe

Hi there, Thanks for keeping up with my questions. As bicycle is able to from differential analysis between two different groups i.e treatment vs control. I wanted to know if bicycle can perform commonly methylated CpG shores, CPG islands, CpG shelves and promoter regions among a group of samples i.e control samples ?

Many thanks

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/sing-group/bicycle/issues/5, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADDOQKYGPSI4RYA64TW77ITRN2XQHANCNFSM4MN7NAEA.

ADVERTENCIA LEGAL: Este correo electrónico, y en su caso los ficheros adjuntos, pueden contener información protegida para el uso exclusivo de su destinatario. Se prohíbe la distribución, reproducción o cualquier otro tipo de transmisión por parte de otra persona que no sea el destinatario. Si usted recibe por error este correo, se ruega comunicarlo al remitente y borrar el mensaje recibido. De conformidad con lo dispuesto en el Reglamento (UE) 2016/679 relativo a la protección de los datos personales de las personas físicas, la información personal que nos pueda facilitar a través de este correo electrónico quedará registrada por la Fundación CNIO con la finalidad de tramitar el objeto del presente correo electrónico. El tratamiento de sus datos personales se encuentra legitimado por ser necesario para gestionar el objeto del presente mensaje. Estos datos personales no serán comunicados a ningún destinatario salvo a aquellos que usted nos autorice o así venga exigido por una ley. Ud. podrá ejercer los derechos de acceso, rectificación, supresión, limitación de tratamiento, portabilidad y oposición en la siguiente dirección: c/Melchor Fernandez Almagro 3, 28029 (Madrid). Podrá ponerse en contacto con el Delegado de Protección de Datos en: delegado_lopd@cnio.es. Para el caso de que Ud. precise conocer información adicional sobre el tratamiento de sus datos personales, puede consultar dicha información adicional en el siguiente enlace dentro de nuestra página web: https://www.cnio.es/es/privacidad/index.asp

LEGAL NOTICE: This email and any attached files may contain protected information for the sole use of its intended recipient or addressee. Anyone other than the intended recipient or addressee is strictly prohibited from distributing, reproducing or transmitting the email and its contents in any way. If you receive this email in error, please notify the sender and delete the message. Pursuant to the provisions of EU Regulation 2016/679 regarding the protection of personal data, any personal information you provide through this email will be registered by the CNIO Foundation in order to deal with content of this email. Your personal data must be processed in order to be able to deal with the content and purpose of this message. Your personal details will not be passed on to anyone else unless you authorise us to do so or we are required to do so by law. You may exercise your rights regarding access, rectification, suppression, limitation of processing, portability and opposition by writing to the following address: c/Melchor Fernandez Almagro 3, 28029 (Madrid). You may contact the Data Protection Delegate (Delegado de Protección de Datos) at: delegado_lopd@cnio.es. If you require further information about the processing of your personal data, go to the following link on our webpage: https://www.cnio.es/es/privacidad/index.asp

[freedomq8]

Thanks for clarification. It a new field for me so wanted to know more about the approach to achieve common methylation across the same group of interest. As you explained that in order to achieve the annotation of such functional regions is to employ BED file annotation. Which I will definitely explore.

Nonetheless, I am still looking for the best way to look for similar methylated region within certain condition i.e control. So what is the best way to achieve this and which bicycle file output I will be using ?

Thanks

[osvaldogc]

Hi again,

Bicycle provides several output files: there are output files with quantified methylation levels per sample and per region provided, and files with differential methylation levels between two conditions. Regarding your question, I believe that you are interested in those files with quantification levels. In case that you provided a CpG sites annotation (for instance), Bicycle will give you methylation levels for each CpG within those annotated regions. It will output a global methylation level during the excecution, but perphaps the best way is to browse that quantification tables and directly select within them those regions that you would like to average (at the end it is a table that you can open with a spreadsheet).

osvaldo.

http://www.cnio.es/

---

From: freedomq8 [notifications@github.com] Sent: Saturday, April 25, 2020 4:20 AM To: sing-group/bicycle Cc: Graña.Osvaldo; Comment Subject: Re: [sing-group/bicycle] Commonly methylated (#5)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe

Thanks for clarification. It a new field for me so wanted to know more about the approach to achieve common methylation across the same group of interest. As you explained that in order to achieve the annotation of such functional regions is to employ BED file annotation. Which I will definitely explore.

Nonetheless, I am still looking for the best way to look for similar methylated region within certain condition i.e control. So what is the best way to achieve this and which bicycle file output I will be using ?

Thanks

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/sing-group/bicycle/issues/5#issuecomment-619305368, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADDOQK3R7ITPMQRF4ABZ7G3ROJCFLANCNFSM4MN7NAEA.

ADVERTENCIA LEGAL: Este correo electrónico, y en su caso los ficheros adjuntos, pueden contener información protegida para el uso exclusivo de su destinatario. Se prohíbe la distribución, reproducción o cualquier otro tipo de transmisión por parte de otra persona que no sea el destinatario. Si usted recibe por error este correo, se ruega comunicarlo al remitente y borrar el mensaje recibido. De conformidad con lo dispuesto en el Reglamento (UE) 2016/679 relativo a la protección de los datos personales de las personas físicas, la información personal que nos pueda facilitar a través de este correo electrónico quedará registrada por la Fundación CNIO con la finalidad de tramitar el objeto del presente correo electrónico. El tratamiento de sus datos personales se encuentra legitimado por ser necesario para gestionar el objeto del presente mensaje. Estos datos personales no serán comunicados a ningún destinatario salvo a aquellos que usted nos autorice o así venga exigido por una ley. Ud. podrá ejercer los derechos de acceso, rectificación, supresión, limitación de tratamiento, portabilidad y oposición en la siguiente dirección: c/Melchor Fernandez Almagro 3, 28029 (Madrid). Podrá ponerse en contacto con el Delegado de Protección de Datos en: delegado_lopd@cnio.es. Para el caso de que Ud. precise conocer información adicional sobre el tratamiento de sus datos personales, puede consultar dicha información adicional en el siguiente enlace dentro de nuestra página web: https://www.cnio.es/es/privacidad/index.asp

LEGAL NOTICE: This email and any attached files may contain protected information for the sole use of its intended recipient or addressee. Anyone other than the intended recipient or addressee is strictly prohibited from distributing, reproducing or transmitting the email and its contents in any way. If you receive this email in error, please notify the sender and delete the message. Pursuant to the provisions of EU Regulation 2016/679 regarding the protection of personal data, any personal information you provide through this email will be registered by the CNIO Foundation in order to deal with content of this email. Your personal data must be processed in order to be able to deal with the content and purpose of this message. Your personal details will not be passed on to anyone else unless you authorise us to do so or we are required to do so by law. You may exercise your rights regarding access, rectification, suppression, limitation of processing, portability and opposition by writing to the following address: c/Melchor Fernandez Almagro 3, 28029 (Madrid). You may contact the Data Protection Delegate (Delegado de Protección de Datos) at: delegado_lopd@cnio.es. If you require further information about the processing of your personal data, go to the following link on our webpage: https://www.cnio.es/es/privacidad/index.asp

[freedomq8]

Dear osvaldo,

Thanks for your reply and patient,

I am still figuring out what would be the best way to find common methylated regions for group of individuals with phenotype of interest. I am also using a bed file supplied with Illumina which is the targeted regions. I also managed to annotate the data from UCSC. Now I am trying to find which column to use and which parameters to use as cutoff (normally) so I can use it and see what is in common across the same condition samples. lets say I am interested in CpG regions, which makes me consider mCG total methylation , mCG total depth, and/or WMCM CG TOTAL right ? if yes which values scientists use as a cutoff. I extract some rows from one sample so I can understand the values in biological terms especially WMCM CG TOTAL? what is the question mark symbol stand for, also what does 0 mean (not methylated, 1 and 0.8).

`

Region | mCG WATSON | depthCG WATSON | WMCM CG WATSON | mCG CRICK | depthCG CRICK | WMCM CG CRICK | mCG total methylation | mCG total depth | WMCM CG TOTAL | mCHG WATSON | depthCHG WATSON | WMCM CHG WATSON | mCHG CRICK | depthCHG CRICK | WMCM CHG CRICK | mCHG total methylation | mCHG total depth | WMCM CHG TOTAL | mCHH WATSON | depthCHH WATSON | WMCM CHH WATSON | mCHH CRICK | depthCHH CRICK | WMCM CHH CRICK | mCHH total methylation | mCHH total depth | WMCM CHH TOTAL

-- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- chr1_796248-796660 | 0 | 0 | � | 0 | 0 | � | 0 | 0 | � | 0 | 2 | 0 | 0 | 0 | � | 0 | 2 | 0 | 0 | 10 | 0 | 0 | 0 | � | 0 | 10 | 0 chr1_811811-812540 | 0 | 0 | � | 0 | 2 | 0 | 0 | 2 | 0 | 0 | 0 | � | 0 | 18 | 0 | 0 | 18 | 0 | 0 | 0 | � | 0 | 46 | 0 | 0 | 46 | 0 chr21_47801759-47804946 | 2714 | 2789 | 0.973109 | 943 | 980 | 0.962245 | 3657 | 3769 | 0.970284 | 27 | 2514 | 0.01074 | 4 | 1438 | 0.002782 | 31 | 3952 | 0.007844 | 61 | 4014 | 0.015197 | 20 | 2696 | 0.007418 | 81 | 6710 | 0.012072 chr15_39490882-39490886 | 0 | 0 | � | 5 | 5 | 1 | 5 | 5 | 1 | 0 | 0 | � | 0 | 5 | 0 | 0 | 5 | 0 | 0 | 0 | � | 0 | 0 | � | 0 | 0 | �

` Thanks for your help.

N.B: I used to different bed files, one the targeted regions by Illumina for methylcytosine calling step and the annotation such as ucsc N_shore islands in the differential step. I noticed the columns shifted and when using the annotation. I also suspect if i only use the annotation file for methylation calling this would give a different results, am i right ?

[osvaldogc]

Good morning !

Sorry, I quickly checked my email and yours fall into my set of read emails, but without having answered it.

How many different phenotypes (groups of samples) do you have?. Two phenotypes? Why don't you perform differential methylation between them and so you could base your selection of regions of interest as those methylated regions with p-adjusted < 0.05 and some log2FC value. Depending on your specific requirements, you could set it as a higher or lower value of log2FC. This will give you a set of methylated regions over a log2FC threshold for the case group (assuming it as the phenotype of interest). Once that you have selected all those cases, you could check the variance in methylation along the samples of the group of interest.

Does this sound reasonable ?

osvaldo.

---

From: freedomq8 [notifications@github.com] Sent: Friday, May 01, 2020 12:29 PM To: sing-group/bicycle Cc: Graña.Osvaldo; Comment Subject: Re: [sing-group/bicycle] Commonly methylated (#5)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe

Dear osvaldo,

Thanks for your reply and patient,

I am still figuring out what would be the best way to find common methylated regions for group of individuals with phenotype of interest. I am also using a bed file supplied with Illumina which is the targeted regions. I also managed to annotate the data from UCSC. Now I am trying to find which column to use and which parameters to use as cutoff (normally) so I can use it and see what is in common across the same condition samples. lets say I am interested in CpG regions, which makes me consider mCG total methylation , mCG total depth, and/or WMCM CG TOTAL right ? if yes which values scientists use as a cutoff. I extract some rows from one sample so I can understand the values in biological terms especially WMCM CG TOTAL? what is the question mark symbol stand for, also what does 0 mean (not methylated, 1 and 0.8.

`

Region mCG WATSON depthCG WATSON WMCM CG WATSON mCG CRICK depthCG CRICK WMCM CG CRICK mCG total methylation mCG total depth WMCM CG TOTAL mCHG WATSON depthCHG WATSON WMCM CHG WATSON mCHG CRICK depthCHG CRICK WMCM CHG CRICK mCHG total methylation mCHG total depth WMCM CHG TOTAL mCHH WATSON depthCHH WATSON WMCM CHH WATSON mCHH CRICK depthCHH CRICK WMCM CHH CRICK mCHH total methylation mCHH total depth WMCM CHH TOTAL

chr1_796248-796660 0 0 � 0 0 � 0 0 � 0 2 0 0 0 � 0 2 0 0 10 0 0 0 � 0 10 0 chr1_811811-812540 0 0 � 0 2 0 0 2 0 0 0 � 0 18 0 0 18 0 0 0 � 0 46 0 0 46 0 chr21_47801759-47804946 2714 2789 0.973109 943 980 0.962245 3657 3769 0.970284 27 2514 0.01074 4 1438 0.002782 31 3952 0.007844 61 4014 0.015197 20 2696 0.007418 81 6710 0.012072 chr15_39490882-39490886 0 0 � 5 5 1 5 5 1 0 0 � 0 5 0 0 5 0 0 0 � 0 0 � 0 0 �

` Thanks for your help.

N.B: I used to different bed files, one the targeted regions by Illumina for methylcytosine calling step and the annotation such as ucsc N_shore islands in the differential step. I noticed the columns shifted and when using the annotation. I also suspect if i only use the annotation file for methylation calling this would give a different results, am i right ?

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/sing-group/bicycle/issues/5#issuecomment-622334480, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADDOQKZUN2QQI2VGC6SH7HDRPKQAVANCNFSM4MN7NAEA.

ADVERTENCIA LEGAL: Este correo electrónico, y en su caso los ficheros adjuntos, pueden contener información protegida para el uso exclusivo de su destinatario. Se prohíbe la distribución, reproducción o cualquier otro tipo de transmisión por parte de otra persona que no sea el destinatario. Si usted recibe por error este correo, se ruega comunicarlo al remitente y borrar el mensaje recibido. De conformidad con lo dispuesto en el Reglamento (UE) 2016/679 relativo a la protección de los datos personales de las personas físicas, la información personal que nos pueda facilitar a través de este correo electrónico quedará registrada por la Fundación CNIO con la finalidad de tramitar el objeto del presente correo electrónico. El tratamiento de sus datos personales se encuentra legitimado por ser necesario para gestionar el objeto del presente mensaje. Estos datos personales no serán comunicados a ningún destinatario salvo a aquellos que usted nos autorice o así venga exigido por una ley. Ud. podrá ejercer los derechos de acceso, rectificación, supresión, limitación de tratamiento, portabilidad y oposición en la siguiente dirección: c/Melchor Fernandez Almagro 3, 28029 (Madrid). Podrá ponerse en contacto con el Delegado de Protección de Datos en: delegado_lopd@cnio.es. Para el caso de que Ud. precise conocer información adicional sobre el tratamiento de sus datos personales, puede consultar dicha información adicional en el siguiente enlace dentro de nuestra página web: https://www.cnio.es/es/privacidad/index.asp

LEGAL NOTICE: This email and any attached files may contain protected information for the sole use of its intended recipient or addressee. Anyone other than the intended recipient or addressee is strictly prohibited from distributing, reproducing or transmitting the email and its contents in any way. If you receive this email in error, please notify the sender and delete the message. Pursuant to the provisions of EU Regulation 2016/679 regarding the protection of personal data, any personal information you provide through this email will be registered by the CNIO Foundation in order to deal with content of this email. Your personal data must be processed in order to be able to deal with the content and purpose of this message. Your personal details will not be passed on to anyone else unless you authorise us to do so or we are required to do so by law. You may exercise your rights regarding access, rectification, suppression, limitation of processing, portability and opposition by writing to the following address: c/Melchor Fernandez Almagro 3, 28029 (Madrid). You may contact the Data Protection Delegate (Delegado de Protección de Datos) at: delegado_lopd@cnio.es. If you require further information about the processing of your personal data, go to the following link on our webpage: https://www.cnio.es/es/privacidad/index.asp

[freedomq8]

Morning to you,

its alright, just happy to do any productivity during this pandemic. its one phenotype, i personally didn't design the experiment so not sure why. Nevertheless, its more than 15 samples of the same phenotype and wanted to see what is commonly methylated among these individuals who have the same/one phenotype. I did not what is the best file to use, so i guess it is the file that (sample).METHYLATEDregions.txt. Now i wanted to use a cut off parameters to priorities the methylated regions for each sample that later will see what are the regions that is common between all samples who have the same and only phenotype (I don't have another phenotype to do differentially analysis). Now doing, as we are interested in CpG methylated regions. i guess we will be interested in CG total methylation , mCG total depth, and/or WMCM CG TOTAL columns (am I right)? if yes, I found a range of values also symbols � and zeros which now i am trying to determine a cut of for each or one of these column to consider this region is methylated but i am not sure which value to use as these are numbers. So this is my question is there a common thershold/cut-off to consider in using (sample).METHYLATEDregions.txt file and G total methylation , mCG total depth, and/or WMCM CG TOTAL columns to prioritise/identify methylated regions ?

so there is no groups just one phenotype.

Hope this help and looking forward to your feedback.

[osvaldogc]

Yes, I would use the METHYLATEDregions.txt file.

As a first approach, in order to have an idea of how the data looks like, I would probably do the following:

(1) Select the set of annotated regions that you are interested in (probably all the regions within the aforementioned file).

(2) For each region in (1) and for each sample, read the WMCM CG TOTAL value, producing a new table with defined regions as rows and and WMCM value of each sample as columns. So you end up with a new table with regions (rows) and WMCM (columns) of the samples.

(3) Now, for each one of those regions calculate the methylation median across the samples and the standard deviation: from this you should get for each annotated region a measure of how consistent is the methylation across the samples (i.e. whether there are similar levels of methylation in all the samples or not).

(4) For the cases in (3) where there is almost no standard deviation from the median you could rescue the median value of methylation to a third table.

(5) From the new table built in (4) calculate the distribution of methylation for all the regions inside it.

(6) Once that you have the samples sorted according to the methylation levels in (5), then you could set a dynamic threshold of methylation (0.9, 0.8, 0.7, ......, etc) and point out the regions that fall within each threshold of methylation. Doing it this way does not require that you set an arbitrary threshold of methylation, because you are checking all the methylation regions that fall under each methylation threshold.

Does it sound reasonable ?

osvaldo.

---

From: freedomq8 [notifications@github.com] Sent: Tuesday, May 05, 2020 1:50 PM To: sing-group/bicycle Cc: Graña.Osvaldo; Comment Subject: Re: [sing-group/bicycle] Commonly methylated (#5)

CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe

Morning to you,

its alright, just happy to do any productivity during this pandemic. its one phenotype, i personally didn't design the experiment so not sure why. Nevertheless, its more than 15 samples of the same phenotype and wanted to see what is commonly methylated among these individuals who have the same/one phenotype. I did not what is the best file to use, so i guess it is the file that (sample).METHYLATEDregions.txt. Now i wanted to use a cut off parameters to priorities the methylated regions for each sample that later will see what are the regions that is common between all samples who have the same and only phenotype (I don't have another phenotype to do differentially analysis). Now doing, as we are interested in CpG methylated regions. i guess we will be interested in CG total methylation , mCG total depth, and/or WMCM CG TOTAL columns (am I right)? if yes, I found a range of values also symbols � and zeros which now i am trying to determine a cut of for each or one of these column to consider this region is methylated but i am not sure which value to use as these are numbers. So this is my question is there a common thershold/cut-off to consider in using (sample).METHYLATEDregions.txt file and G total methylation , mCG total depth, and/or WMCM CG TOTAL columns to prioritise/identify methylated regions ?

so there is no groups just one phenotype.

Hope this help and looking forward to your feedback.

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/sing-group/bicycle/issues/5#issuecomment-624007942, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADDOQK6YBQCDYF5BQNUQHMDRP74ORANCNFSM4MN7NAEA.

ADVERTENCIA LEGAL: Este correo electrónico, y en su caso los ficheros adjuntos, pueden contener información protegida para el uso exclusivo de su destinatario. Se prohíbe la distribución, reproducción o cualquier otro tipo de transmisión por parte de otra persona que no sea el destinatario. Si usted recibe por error este correo, se ruega comunicarlo al remitente y borrar el mensaje recibido. De conformidad con lo dispuesto en el Reglamento (UE) 2016/679 relativo a la protección de los datos personales de las personas físicas, la información personal que nos pueda facilitar a través de este correo electrónico quedará registrada por la Fundación CNIO con la finalidad de tramitar el objeto del presente correo electrónico. El tratamiento de sus datos personales se encuentra legitimado por ser necesario para gestionar el objeto del presente mensaje. Estos datos personales no serán comunicados a ningún destinatario salvo a aquellos que usted nos autorice o así venga exigido por una ley. Ud. podrá ejercer los derechos de acceso, rectificación, supresión, limitación de tratamiento, portabilidad y oposición en la siguiente dirección: c/Melchor Fernandez Almagro 3, 28029 (Madrid). Podrá ponerse en contacto con el Delegado de Protección de Datos en: delegado_lopd@cnio.es. Para el caso de que Ud. precise conocer información adicional sobre el tratamiento de sus datos personales, puede consultar dicha información adicional en el siguiente enlace dentro de nuestra página web: https://www.cnio.es/es/privacidad/index.asp

LEGAL NOTICE: This email and any attached files may contain protected information for the sole use of its intended recipient or addressee. Anyone other than the intended recipient or addressee is strictly prohibited from distributing, reproducing or transmitting the email and its contents in any way. If you receive this email in error, please notify the sender and delete the message. Pursuant to the provisions of EU Regulation 2016/679 regarding the protection of personal data, any personal information you provide through this email will be registered by the CNIO Foundation in order to deal with content of this email. Your personal data must be processed in order to be able to deal with the content and purpose of this message. Your personal details will not be passed on to anyone else unless you authorise us to do so or we are required to do so by law. You may exercise your rights regarding access, rectification, suppression, limitation of processing, portability and opposition by writing to the following address: c/Melchor Fernandez Almagro 3, 28029 (Madrid). You may contact the Data Protection Delegate (Delegado de Protección de Datos) at: delegado_lopd@cnio.es. If you require further information about the processing of your personal data, go to the following link on our webpage: https://www.cnio.es/es/privacidad/index.asp

[freedomq8]

Thanks a lot, it does make sense and will work on it. Mean while can I know what does this symbol mean in the WMCM column � ? is it unknown ? also should I consider another column for sitting a threshold apart from WMCM or is using it will be enough?

thanks for your patience