Overview

The UGlib profiler is designed so that it can be initiated with libraries from a focused set of tissues (as in the case of the Mouse Mammary/MammaryGland) or can be used to access libraries from all tissue classifications (as in the case of All Mouse or Human Libraries).

The UGlib Profiler tool is used to screen UniGene clusters for ones whose "EST members" meet certain criteria which you specify with the tool. You select specific libraries into one or two Groups (Group "A" or Group "B").

You must select at least one library into Group "A". This is done by clicking on the "radio" in the "A" column to the left of the library. You can select more than one library into Group "A". If no libraries are selected into Group "B", the results returned will be all clusters whose membership includes at least one sequence from a Group "A" library. You can exclude singletons (clusters of size one) from the results by checking the "Exclude Singletons" box. To find Clusters unique to Group "A" check the "NOT in all others" box.

You can also place libraries in Group "B". You can then screen for clusters with members from BOTH Group "A" and Group "B" (check the "and" radio button) or with members from Group "A" but NOT Group "B" (check the "not" radio button).

Links from the UGlib Profiler Query form and the returned Preview results:

• The link on the "lib.xxx" will take you to the designated NCBI UniGene library.
• The link on Hs.xxxx or Mm.xxxx will take you to the designated NCBI UniGene cluster page.
• The link on the "TP" (to the left of the UniGene Cluster) will generater a tissue profile page for that cluster.
  Note of Caution: The "Normalized" values displayed on the Tissue Profile are based strictly on equalizing the "total sample size" for each tissue represented in the cluster. See things to consider at the end of this page.

Examples

Things to consider when looking at sequence derived UniGene "expression profiles":

1. UniGene is a clustering of EST sequences. The clustering is not perfect. The sequence data is noisy (it is filtered). Clustering is based on a 50 basepair window.
2. Some clones have more than one sequence (for example they may have been sequenced from both ends). If the sequences are in the same cluster, there will be multible counts for one clone.
3. There are non-normalized libraries, normalized libraries, subtracted libraries, pooled libraries etc. in dbEST. This is not taken into consideration in the analysis.
4. We rely on the UniGene Tissue classification, some libraries may not yet be classified or may be missclassified.
5. There are large differences in sample sizes between libraries and tissue classifications

mAdb Home Page | mAdb Gateway | Upload Status Reference Info | Program Downloads | GeneCards
	NIH BioInformatics support provided by BIMAS/CBEL/CIT. We can be contacted by email.