Subtractive assembly for comparative metagenomics

• Introduction: Subtractive Assembly is a de novo approach for comparative metagenomics by assembling the differential portion of reads directly. Given two metagenomic datasets A and B, substractive assembly approach first identifies reads that are likely sampled from abundant or unique genomes in A (or B vice versa), using Bloom-filter based k-mer counting. And one can assemble only these reads separately using metagenomic assemblers such as IDBA-UD.
• Download the package from SourceForge
• Get the package with example datasets for using the SAcounter: SAtest.tz (Warning: big file!)
• The subtractive assembly results of the type 2 diabetes (T2D) metagenomes:
  • The set of genes identified uniquely by subtractive assembly in the T2D cohort:T2D_new_gene.faa
  • Annotations for the genes above based on SEED classification:functions.tbl and meta_recon.tbl
  • The set of genes above that passed Wilcoxon rank-sum test corrected by false discovery rate (with q-value cutoff 0.01) in the T2D cohort:T2D_new_gene_fdr.faa
  • Annotations for the genes that passed t-test based on SEED classification:functions_fdr.tbl and meta_recon_fdr.tbl
• Citation: Mingjie Wang, Thomas G. Doak, and Yuzhen Ye. Subtractive assembly for comparative metagenomics, and its application to type 2 diabetes metagenomes. Genome Biology, 2015 (accepted)
• Contact: Yuzhen Ye (yye@indiana.edu) (web)