AraCyc was developed by TAIR for the annotation of biochemical pathways in Arabidopsis thaliana.

Gene expression data used in ATTED-II are collected by AtGenExpress consortium, and are download from TAIR.

Most gene pairs show random-level coexpression scores, which are normally distributed. Therefore, we adopted a coexpression z-score, where 0 is a random level, and 3 is a false positive rate of 0.1% for any species or version. Some gene pairs show larger coexpressions than the normal distribution of the random coexpression. The deviation is apparent around z>3, suggesting that coexpression with z>3 is a useful when using ATTED-II.

Five AtGenExpress subgroups were used to construct condition-specific coexpression. [more].

• Tissue (237)
• Abiotic stress (534)
• Biotic stress (200)
• Hormone (200)
• Light (48)

The annotation file ("gene_info") released by NCBI is used in ATTED-II.

Gene Ontology released by TAIR was used.
Prefix of original IDs of GO were changed in ATTED-II according to those type (BP, CC and MF).

KEGG (Kyoto Encyclopedia of Genes and Genemes) was developed by Genomenet. Annotation of biochemical pathways were used in ATTED-II.

Logit Score between a pair of genes is calculated as follows,

• p = MR / N
• logit(p) = log2(p / (1-p))
• LS = - logit(p)
• N: Number of genes in the platform.

About logit function in coexpression calculation, please also check our paper.
Obayashi et al. (2018) ATTED-II in 2018: A Plant Coexpression Database Based on Investigation of the Statistical Property of the Mutual Rank Index.

Mutual rank is calculated for edge strength of coexpressed gene network. [more]

Annotation released by NCBI Gene.
Download/FTP : https://ftp.ncbi.nih.gov/gene/DATA/ASN_BINARY/Plants/All_Plants.ags.gz

Sample contribution score is calculated as a product of z-scored expression values. Avarage of the score is pearsons correlation coefficient.

Two tools are used to predict localizations of proteins; TargetP and WoLF PSORT. In t\ ables in ATTED-II, long target names pop up when the mouse is over the abbrevations.

TargetP results were released by TAIR. c:m:s:o scores represent reliabilities for chlor\ oplast, mitochondoria, scretory, and others.
Stand alone program of WoLF PSORT is provided by Dr. Paul Horton in Computational Biology Research Center, AIST, Japan.

Result of TargetP released by TAIR
c:m:s:o scores represent reliabilities for chloroplast, mitochondoria, scretory, and others.

The following species uses multiple taxonomy IDs for different purposes.
- Ppo:
 - ncbi gene: 3694, 3697
 - RNAseq coexpression
  - SRA retrieval: 3694
  - SRA mapping (refseq): 3697

Result of WoLF PSORT calculated by Dr. Paul Horton in Computational Biology Research Center, AIST, Japan.