Liverome is a curated database of liver cancer-related gene signatures. The gene signatures were obtained mostly from published microarray and proteomic studies, and thoroughly curated by experts.

Reference: Lee L, Wang K, Li G, Xie Z, Wang Y, Xu J, Sun S, Pocalyko D, Bhak J, Kim C, Lee KH, Jang YJ, Yeom YI, Yoo HS, Hwang S. (2011) Liverome: a curated database of liver cancer-related gene signatures with self-contained context information. BMC Genomics, 12(Suppl 3):S3. PubMed PDF

Presentation slides on Liverome at International Conference on Bioinformatics 2011: PDF

To search on a single gene

Use Gene search menu to retrieve all available differential expression information on a gene.

To search on multiple genes (i.e., a gene list)

• To retrieve all available differential expression information on multiple genes:
  1. Download all collected gene lists (zipped, 8.7MB)
  2. Unzip and look for an Excel file named combined_gene_list_file.xls. It lists all the genes that have at least one published evidence of differential expression.
  3. To the Excel file, append a new column containing the genes in your list as either official gene symbol or Entrez Gene ID. Upon appropriate manual sorting and comparison within Excel, you should be able to obtain all available differential expression information on your genes.
• To compare your gene list with several lists from the signature collection:
  1. Go to View/Compare Lists menu.
  2. Paste your gene list, select several of the collected lists, and compare them.

1. For every gene signature, compared sample groups are explictly designated

• Since differential expression experiment involves comparison between sample groups, measures of differential expression (e.g., fold change) of a gene is meaningful only when we know the nature of the compared sample groups (e.g., recurrence group vs. non-recurrence group; tumor vs. normal; HBV-tumor vs. HCV-tumor).
• Nevertheless, in most other gene signature databases, this important piece of information is missing or ambiguous since it can be prepared only by tedious manual curation.
• Therefore, for every gene signature, we manually curated information on the compared sample groups to make self-contained database content.

2. For every gene signature, summary of the experiment is prepared

• A summary of experiment that produced the gene signature would be helpful to quickly characteristics of the gene signature (e.g., platform, sample size, study design, clinicopathological sample characteristics).
• Nevertheless, in all other gene signature databases, this important piece of information is missing. Instead, other databases merely copy-and-pasted the Abstract of the corresponding publication.
• Therefore, for every gene signature, we manually curated the summary of the experiment to make self-contained database content.

3. The database content is easy to browse

1. Every gene signature is named informatively.
• For quick recognition of the nature of gene signatures, they were informatively named.
• An example from other database:        Viral_Okamoto06_36genes
• An example from Liverome database: Okamoto (2006) Ann Surg Oncol [Predictive marker genes for multicentric hepatocarcinogenesis]
2. The gene signature collection is meaningfully organized according to functional categories.

4. Largest collection of liver cancer-related gene signatures