##RDP Alignment tools package

The Alignment tools package contains commands to do defined community analysis (Fish et al., 2013. FunGene: the functional gene pipeline and repository. Frontiers, Accepted), pairwise alignment and hidden markov model (HMMER3 models, no training).

This project depends on https://github.com/rdpstaff/ReadSeq and https://github.com/rdpstaff/fungene_pipeline. See RDPTools (https://github.com/rdpstaff/RDPTools) to install.

• Defined Community Analysis

  This command compares input nucleotide reads to the set of known sequences for amplification targets in the sequenced DNA. It computes a global alignment between the read and each of the reference sequences. The pairwise alignment producing the highest alignment score is used to identify the source organism, substitutions, indels and associated quality scores. These output files can be processed by the parseErrorAnalysis.py script (https://github.com/rdpstaff/fungene_pipeline) to summarize the types of errors found in the input reads. The tab-delimited summary file can be imported to Excel programs to make graphs.

    java -jar /path/to/AlignmentTools.jar compare-error-type
    usage: CompareErrorType [options] <ref_nucl> (<query_nucl> | <query_nucl.fasta> <query_nucl.qual>)
    		-s,--stem <arg>   Output stem (default <query_nucl.fasta>)
  

  The sample input and output files can be downloaded from RDP tutorial http://rdp.cme.msu.edu/tutorials/defined_community/RDPtutorial_DEFINED-COM.html. Example commands:

    java -jar /path/to/AlignmentTools.jar compare-error-type -s mid01 /path/to/nifH_control_refseq_nucl_slice.fa  /path/to/mid01_trimmed.fasta /path/to/mid01_trimmed.qual
    		
    /path/to/fungene_pipeline/parseErrorAnalysis.py -i chimera_contaminant.id -q mid01_qual.txt mid01_alignments.txt mid01_mismatches.txt mid01_indels.txt /path/to/nifH_control_refseq_nucl_slice.fa> mid01_erroranalysis_summary.txt
  

• Merge Alignment This command merges multiple alignment files into one file. The input files can be in fasta or stockholm format, each with a reference sequence named "#=GC_RF" to mark model positions.

    java -jar /path/to/AlignmentTools.jar alignment-merger alignment merged_aligned.fasta
  

• Compute k-nearest-neighbors by pairwise alignment

  This command computes pairwise alignment between the read and each reference sequences. It returns the alignments with the top k highest scores. It reports the following for each alignment: seqname k ref_seqid ref_desc orientation score ident query_start query_end query_length ref_start ref_end

    java -jar /path/to/AlignmentTools.jar pairwise-knn         
    usage: PairwiseKNN <options> <queryFile> <dbFile>
    -k <arg>          K-nearest neighbors to return
    -m,--mode <arg>   Alignment mode {global, glocal, local, overlap,
               overlap_trimmed} (default= glocal)
    -o,--out <arg>    Redirect output to file instead of stdout