Manual curation of metadata have 3 levels : Sample, SubProject and Project.

The structure of curation model is as follows:

It systematically integrates 179,079 high-quality cells, involving 16 projects and 5 strategy from NCBI, EBI, DDBJ and NGDC (GEN). It also enables Browse, Search, Analysis, Visualization and Download.

Some of the attributes are listed here:

Same as RNA-seq
'LibraryStrategy'='10x genomics' or
'LibraryStrategy'='Smart-seq2' or
'LibraryStrategy'='SMARTer' or
'LibraryStrategy'='Smart-seq v4' or
'LibraryStrategy'=' Drop-seq'

scRNA-seq pipeline is constructed with reference to GEN toolkits (https://ngdc.cncb.ac.cn/gen/documentation).It includes quality control, read alignments, gene/transcript expression quantification and cell clustering. 

10X:

1.Extract barcode, UMI, RNA read

2.Correct barcode

3.Aligned reads by STAR

4.Tag reads with genes, transcript hits

5.Count UMIs

6.Select cell barcodes

Drop-seq:

1.Drop tag

2.Extract barcode, UMI, RNA read

3.Aligned reads by STAR

4.Generate dropEST and dropReport

SMART-seq:

1.Filter low quality reads: Fastp v0.20.0 (Chen et al, 2018)

2.Strandness library: RseQC v2.6.4 (Wang et al, 2012)

3.Mapping to reference genome: STAR 2.7.1a(Dobin et al, 2013) 

4.Gene/isoform assembly and quantification: RSEM v1.3.1 (Li & Dewey, 2011)

5.Basic expression profiling

The same as 2.4 Meta Curation

It systematically integrates 26 high-quality methylation samples, involving 9 projects. It also enables Browse, Search, Analysis, Visualization and Download. Besides, it provides manually curated knowledge of both featured differentially methylated genes (DMGs) across 12 kinds of biological contexts like disease, as well as methylation tools collection.

Some of the attributes are listed here:
The status of data resource is open-access

'DataSet Type'= 'methylation profiling by high throughput sequencing'

[All Fields] = "WGBS" or "BS-Seq" or "Whole Genome Bisulfite Sequencing"

The predicted sequencing depth of WGBS sample should be greater than 10

WGBS-seq pipeline is constructed with reference to MethBank toolkits

Quality control: FastQC v0.11.7, Fastq-dump (sratoolkit.2.8.2-1)

Mapping to the reference genome: Bismark-0.22.3

Visualizing: Mapping rate, Unique mapping rate, Genome coverage, C coverage, Conversion rate and Depth.

Manual curation of metadata have 2 levels: Sample and Project.

The structure of curation model is as follows:

To further explore human disease and health, we manually curated 1229 NHP model animal disease articles encompassing 308 diseases from 21 disease ontology systems recorded in disease model. Users in any biomedical field can apply the Disease Module to browse disease name or DOID of interest accessing omics data, literature detail information and research situation in other model animal databases.

In the “Disease” part, we present all the diseases related to human health and link disease nodes directly to the curated literatures. Users can conveniently browser the specific disease and find corresponding research articles.
The network presented in the top-left corner is analyzed by Cytoscape. It shows MCL clustered subnetworks from the Disease Ontology (DO) tree which includes all the disease nodes from our curations. Each color block represents a type of systematic disease which is labeled and bolded on the map. The small colored points inside each subnetwork reveal the number of curations. White means zero curation. Blue means 1-5 curations. Orange means 6-14 curations. Purple means over 15 curations. After clicking the subnetwork (i.e. the color block), details will appear on the lower region in the webpage. Users could click the nodes of details to explore curations which showed in the right part of the webpage. We provide some basic information for browsing. More info can be obtained by clicking the PubMed ID.

First, all the Disease Ontology (DO) IDs and their relationships were downloaded from the DO website (https://disease-ontology.org). Then, we mapped our curated DO terms to the downloaded DO tree and extracted the basic necessarily nodes and relations (“Is a”) by customized python and R scripts. The basic DO structures were loaded into Cytoscape (V3.9.1) network analyzing tool and annotated by curated information. A new network was generated and simplified by removing some high-level terms such as Disease (DOID:4), Nervous system disease (DOID:863) and Disease of anatomical entity (DOID:7). Then the new network was created by clustering with the MCL clustering algorithm from the clusterMaker2 (v1.2.1) Cytoscape app. The layout of the new network was manually edited to orderly arrange these clusters.

First, all the Disease Ontology (DO) IDs and their relationships were downloaded from the DO website (https://disease-ontology.org). Then, we mapped our curated DO terms to the downloaded DO tree and extracted the basic necessarily nodes and relations (“Is a”) by customized python and R scripts. The basic DO structures were loaded into Cytoscape (V3.9.1) network analyzing tool and annotated by curated information. A new network was generated and simplified by removing some high-level terms such as Disease (DOID:4), Nervous system disease (DOID:863) and Disease of anatomical entity (DOID:7). Then the new network was created by clustering with the MCL clustering algorithm from the clusterMaker2 (v1.2.1) Cytoscape app. The layout of the new network was manually edited to orderly arrange these clusters.

The structure of curation model is as follows:

Firstly, users can find more details of gene information on this page. The gene card shows gene symbol name, gene description, species, gene location, type of gene and HGNC ID, which was collected from Ensembl BioMart. Besides, users can also query gene when choosing target species and target gene. Multi-species and multi-omics information is reunited into the Gene page, and users can query Transcripts, Gene Ontology, Homology gene, transcriptome and methylome expression information and JBrowse visualization information.

Tools pages display manually curated NHP related tools and softwares for users’ access and query.

nhp_atlas@big.ac.cn

Postal Address:
National Genomics Data Center
China National Center for Bioinformation / Beijing Institute of Genomics
Chinese Academy of Sciences
No.1 Beichen West Road
Chaoyang District, Beijing 100101
China