Colon Cancer - microRNA Expression Database CC-MED

S-MED (Sarcoma Expression Database) F... S-MED Frequently Asked Questions

Q- Why are some datapoints blank?

A- If a blank space is shown in a heatmap or in a data table, a value is not present in the database for the miRNA.  This can occur for a few different reasons. One explanation is that many additional probes were included in phase 2,3,4 that were not present in phase 1. Additionally, a single miRNA probe may in later versions have been split into two different probes for example inphase 1 hsa-miR-335 is present in phase 2 this is now called hsa-miR-335-3p and hsa-miR-335-5p.  hsa-miR-335 would appear missing in experimental phase 2,3,4and hsa-miR-335-3p and -5p would appear missing in experimental phase 1.

Q- What is included in the S-MED Database?
A- Currently the database hosts data from four different experimental phases. Phase 1 consists of 739 miRNAs and 89 tissues. Phase 2 consists of 1134 miRNAs and 92 tissues. Phase 3 consists of 1134 miRNAs and 95 tissue. Phase 4 consists of 1134 miRNAs and 54 tissues .  Thus on the whole, the repository consists of 25 tissue types representing 330 samples. There are over 338000 data-points currently in the dataset.

Q- Why do some miRNA names end in :9.1?
A- Some miRNA names end in 9.1 becasue they are based off build 9.1 of the miRNA database and this changed in 9.1 relative to previous values.

Q-Some of the miRNA sequences can be found at multiple locations with different names why can I find only only one used here?
A-We needed a unique name for each probe so for the 7 probes that had multilpe names in phase 2,3,4 the first miRNA name was used.

Q- How is the data normalized?
A- Each phase was seperatey quantile normalized for basic search.  For advanced search, all data is quantile normalized together.

Q- What statistics are used to generate the the statistics view?
A- The results shown in the statistics view are the results provided for a two group T-test comparing the stated tissue with the remaining tissues found in each phase.  For example the Osteosarcoma tissue results show the results of comparing Osteosarcoma verses the combination of GIST, MFH, FIbromatosis, and Synovial Sarcoma.

Q- Is their a potential for batch affects to be interpreted as differential expression in the advanced search results?
A- Yes, and the user is hereby warned.

Q- I don't see the differential expression of my favorite miRNA eg. hsa-mIR-21 changing, why is that?
A-  Read our paper in BMC cancer. 

Q- I have an awesome suggestion for your database. Who do I contact?
A- Great! send it to sarver at umn dot edu

Q- I have some data I would like to include in your database what should I do?
A- Contact us and we will consider it on a case by case basis.

Q - Who should I contact regarding the contents of this Database?
A-  Dr. Subbaya Subramanian
Assistant Professor
Department of Lab Medicine & Pathology
University of Minnesota
14-271 Moos Tower (Mail code: MMC 206)
515 Delaware St, S.E
Minneapolis, MN-55455
email:subree AT umn DOT edu
Tel: 612-626-4330


@2010-2012 Subramanian Lab  Masonic Cancer Center Biostatistics and Informatics  University of Minnesota
website designed by Aaron Sarver Ph.D.  Hosted by the Minnesota Supercomputing Institute