TransAtlasDB

TransAtlasDB accepts input data from the different software required for differential expression analysis.
The current prototype accepts outputs from the Tuxedo Suite - TopHat and Cufflinks.

Thus, the following information is required to use the TransAtlasDB database application;

1. Sample Information.

The Sample information, or samples metadata, is the reference point of the corresponding results from RNAseq data and therefore important for data archival and retrieval of the various transcriptome analysis. TransAtlasDB preferably accepts the sample information through the FAANG sample submission spreadsheet template to BioSamples. Details are provided here.
The FAANG sample submission spreadsheet template provides a detailed questionnaire for each sample and hence our database system was modeled to accept the FAANG excel template. However, the required fields in the spreadsheet are the Animal and Specimen sheets; with the Animal-‘Sample Name’, Animal-‘Organism’, Specimen-‘Sample Name’ and Specimen-‘Organism Part’ column filled. Also, the Specimen-‘Derived from’ should be the same as the Animal-‘Sample Name’ column of each sample.

The database system also accepts a tab-delimited file with the minimum required columns shown below:

Header	Status	Description
Sample name	required	Sample identification number
Sample description	optional	Description of sample
Derived from	required	Animal identification number
Organism	required	Species name
Organism part	required	Name of tissue
First name	optional	Person's first name
Middle initial	optional	Person's middle initial
Last name	optional	Person's last name
Organization	optional	Institutional name

To download the FAANG BioSamples template click here.
To download the tab-delimited template click here.
Otherwise, the sample information can be manually inserted using SQL insert statements.

2. Alignment Information.
The Alignment information is comprised of the alignment files generated from either RNAseq read mappers. Depending on the RNAseq read mapper used, the expected files are;

1. BAM file ( suffix: .bam )
2. Summary file ( i.e. align_summary.txt or Log.final.out )
NOTE: HISAT2 alignment summary, produced as stardard output at default, must be saved as 'align_summary.txt' file
3. Bed files ( i.e. deletions.bed , insertions.bed , junctions.bed )
NOTE: The bed files are optional.

3. Expression Information.
The Expression information consists of the genes FPKM (Fragments Per Kilobase of transcript per million) file generated using Cufflinks or StringTie, and genomic features read counts using various feature summarization tools like HTSeq-count and featurecount. The required files are;

1. StringTie / Cufflinks
• Transcripts GTF ( suffix: .gtf )
NOTE: The transcripts details should be saved with the suffix '.gtf'
2. Kallisto : Transcripts TSV file ( suffix: .tsv ), or
3. Salmon : Transcripts SF file ( suffix: .sf ), or
4. Feature summarization
• htseq-count / featureCounts ( suffix: .counts )
NOTE: The read counts should be saved with the suffix '.counts'
• STAR ( suffix: ReadsPerGene.out.tab )

4. Variant Information.
The Variant information includes the variant VCF file ( suffix: .vcf ) from variant callers, such as GATK, SAMtools (BCFtools).
Optionally, the functional annotations of variants predicted by different bioinformatics tools can also be provided in a tab-delimited format. TransAtlasDB currently accepts functional annotations from two annotation software;
• Ensembl Variant Effect Predictor (VEP) : The output file should be named with suffix .vep.txt.
• ANNOVAR : The default output file with suffix multianno.txt will be imported.

5. FASTQ details.
This is an optional feature to store the sequence files information if previously unavailable. This requires the FASTQC zipped folder (suffix: .zip ).

The input data files accepted via the TransAtlasDB toolkit are shown below. All input data files should be stored in a single folder, the folder name must be the sample name as denoted in the metadata.

TransAtlasDB outputs user-defined queries as a tab-delimited table.

This table is the default output format which is accepted by most text editors or statistics tools such as Microsoft Excel, R and JMP software.
Asides from the tab-delimited format for exporting results, the variant information can be generated as a VCF output using the -vcf flag.

Variant Call Format

Predicted functional annotations and sample metadata are added in the INFO field of the VCF file, using the key “CSQ” and “MTD” respectively.
Data fields are encoded separated by “|”; the order of fields is written in the VCF header.
VCFs produced by TransAtlasDB follow the standard VCF version 4 file format, and can be used for further downstream analysis or visualization using various variant viewers such as the University of California Santa Cruz (UCSC) Genome Browser, JBrowse, Integrative Genomics Viewer (IGV), and other programs that accept VCF files.