Survival analysis //data processing details

39 views
Skip to first unread message

Bianca....@med.uni-duesseldorf.de

unread,
Oct 19, 2021, 8:59:40 AM10/19/21
to cbiop...@googlegroups.com, nikolas.s...@uni-duesseldorf.de, Georg....@med.uni-duesseldorf.de, Rui....@med.uni-duesseldorf.de, Christop...@med.uni-duesseldorf.de

Dear cBioPortal team,


I am a postdoc at the cancer research lab at University Hospital Duesseldorf. 


I am using your survival analysis tool to test the impact the expression of 7-gene-group. Trying to dig deeper, we are puzzling how exactly the survival outcome is determine in your pipeline. 


We are setting the expression cutoff >=2SD for the group of 7 genes.

Is that cut off based on mean or median?


When we are testing those 7 genes as a group, would it be enough if just one gene's expression is above 2SD to be count in the "altered group", correct?


Is there any code you could provide for us to follow the process? We try to retrace how the results are generated to interpret them better.


May you have there already some contents that you can forward to us?


Thank you for your efforts,


Bianca Behrens 



Dr. rer. nat. Bianca Behrens

Universitätsklinikum Düsseldorf
Experimentelle Chirurgische Onkologie
Geb. 12.41.01 R64
Moorenstr 5
40255 Düsseldorf

0211 81-04502




Universitätsklinikum Düsseldorf (UKD) - Anstalt des öffentlichen Rechts, Moorenstraße 5 | 40225 Düsseldorf | Telefon: +49 211 81-00 | Telefax: +49 211 81-04855; Internet: www.uniklinik-duesseldorf.de | E-Mail: uniklinik...@uniklinik-duesseldorf.de

Aufsichtsratsvorsitzender: Prof. Dr. Peter Dominiak
Vorstand: Prof. Dr. Klaus Höffken (Ärztlicher Direktor, Vorstandsvorsitzender) | Ekkehard Zimmer (Kaufmännischer Direktor)
Prof. Dr. Benedikt Pannen | Torsten Rantzsch | Prof. Dr. Joachim Windolf
Steuernummer: 106-5773-0045
Bankverbindung: Stadtsparkasse Düsseldorf | BLZ: 300 501 10 | Kto.-Nr.: 10 001 550 | IK: 26051 00 18  SWIFT-BIC: DUSSDEDDXXX | IBAN: DE 94 3005 0110 0010 001550

Disclaimer: Die in dieser E-Mail und den dazugehörigen Anhängen (zusammen die „Nachricht“) enthaltenen Informationen sind nur für den Adressaten bestimmt und können vertrauliche und/oder rechtlich geschützte Informationen enthalten. Sollten Sie die Nachricht irrtümlich erhalten haben, löschen Sie die Nachricht bitte und benachrichtigen Sie den Absender, ohne die Nachricht zu kopieren oder zu verteilen oder ihren Inhalt an andere Personen weiterzuleiten.

Higgins, David

unread,
Oct 20, 2021, 9:59:01 AM10/20/21
to Bianca....@med.uni-duesseldorf.de, cbiop...@googlegroups.com, nikolas.s...@uni-duesseldorf.de, Georg....@med.uni-duesseldorf.de, Rui....@med.uni-duesseldorf.de, Christop...@med.uni-duesseldorf.de

Hi Bianca,

 

Thank you for reaching out.

 

We are setting the expression cutoff >=2SD for the group of 7 genes.

Is that cut off based on mean or median?

 

The cut off is based on the number of standard deviations from the mean.

 

When we are testing those 7 genes as a group, would it be enough if just one gene's expression is above 2SD to be count in the "altered group", correct?

 

Yes, the “altered group” is any patient or sample in which at least one of the seven genes in your query has the higher expression.

 

Is there any code you could provide for us to follow the process? We try to retrace how the results are generated to interpret them better. May you have there already some contents that you can forward to us?

 

Our Onco Query Language page has some example coding for queries here, including gene expression: https://www.cbioportal.org/oql

 

We have some slides that walk through the use of OQL as well: https://www.cbioportal.org/tutorials#oql

 

We also have a video tutorial that looks at expression data. This particular example linked here uses OQL to compare survival rates: https://youtu.be/YUxVv6pkxD4?t=1442

 

Let us know if you have any further questions,

 

David M. Higgins, Ph.D. | (he/him)

Scientific Community Program Manager

Center for Data-Driven Discovery in Biomedicine (D3b)

Children’s Hospital of Philadelphia, USA

 

 

--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cbioportal/6892e6af14ee42bcb2973d4d0892898f%40med.uni-duesseldorf.de.
** This email originated from an EXTERNAL sender to CHOP. Proceed with caution when replying, opening attachments, or clicking links. Do not disclose your CHOP credentials, employee information, or protected health information to a potential hacker**.

Reply all
Reply to author
Forward
0 new messages