RNA-Bloom v2.0.1

general changes

• work around a bug in ntcard (fixes #43)
• bypass non-ACGTU characters in input reads (fixes #49)

long-read assembly changes

• improvements to assembly sensitivity
• fix bug in processing reads names containing / (fixes #45)
• standardize format of assembled transcript names (fixes #46)
• add more descriptive error messages (fixes #47)
• add warning for assembly with only short transcripts (fixes #48)

RNA-Bloom v2.0.0

general changes

• display error message when there are too few input reads for assembly (fixes #42)
• updated readme on description of output files (fixes #36)
• -ntcard is turned on automatically if -nk is not specified and ntcard is found in your environment

short-read assembly changes

• release unused Bloom filters during redundancy reduction, lowering memory usage (fixes #40)

long-read assembly changes

• major improvements to assembly quality, peak memory, and runtime
• use strobemers instead of k-mers for read subsampling
  • (default) input arguments to -lrsub are changed
• default value of -lrrd is raised from 2 to 3
• remove sequences with no read support (fixes #18)
• fix bug in handling assembly of a single sequence (fixes #44)

RNA-Bloom v1.4.3

general changes

• upgrade to Java 11
• fix critical bug where file reading takes forever to terminate on certain machines (fixes #15)
• functionality of -Q option is changed to filter reads with low average base quality during the assembly stages
  • now support both short reads and long reads
• various bugfixes, optimization, and changes in default parameters/thresholds throughout

short-read assembly changes

• new options for specifying single-end forward/reverse reads: -sef and -ser
  • can be used in conjunction with -left, -right, and -long
  • support for single-end reads in -pool will be implemented in a future release
• max tip length is now set automatically based on sampled read lengths by default
  • can still be set manually with -tiplength option
• transcripts.short.fa is not used anymore during redundancy reduction to reduce computing requirements (fixes #13)
  • only transcripts.fa is used and only transcripts.nr.fa will be generated

long-read assembly changes

• overhaul of long read assembly workflow
• option to polish long reads with short reads (see README for details)
• new option to turn on/off read sub-sampling during long read assembly: -lrsub (default: on)

RNA-Bloom v1.3.1

general changes

• prints error message when options -left,-right,-long are used with -pool

short reads assembly changes

• initial support for single-end reads (see README for usage)

long reads assembly changes

• initial support for PacBio reads; added new option:
  • -lrpb: use PacBio options for minimap2

RNA-Bloom v1.3.0

general changes

• fixed error when Bloom filter size is adjusted due to high false positive rate

short reads assembly changes

• reduced misassemblies

nanopore reads assembly changes

• significantly reduced run time
• significantly reduced number of temporary files and their disk space usage
• low-complexity reads are discarded during error correction
• improved error correction in polyA tails
• potentially chimeric reads are discarded
• removed read clustering and related options
• added new options:
  • -lrop: min proportion of matching bases in read overlaps
  • -lrrd: min read depth required for assembly

RNA-Bloom v1.2.3

general changes

• support installation with Bioconda (see README for details)

short reads assembly changes

• use binary representation for assembled fragments (reducing their file size)
• significantly reduced runtime for transcript assembly stage
• reduced run time and memory usage of overlap assembly algorithm

nanopore reads assembly changes

• use symbolic links instead of copying files
• reduced run time and memory usage of overlap assembly algorithm
• fixed StackOverflowError

RNA-Bloom v1.2.2

general changes

• please update your ntCard to v1.2.1, which supports uracil and has other improvements and bugfixes

short reads assembly changes

• changed Bloom filter architecture for paired k-mers
• multi-threaded graph construction for transcript assembly
• fixed out-of-bound error

nanopore reads assembly changes

• adjusted read clustering algorithm
• new options to adjust read clustering:
  • -hpc: use homopolymer-compressed minimizers
  • -m: minimizer size
  • -mw: minimizer window size
  • -sop: minimum proportion of sketch overlap minimizers
  • -son: minimum number of sketch overlap minimizers
• reduced memory usage for assembly stage

RNA-Bloom v1.2.1

general changes

• Optional syntax for the -k option in finding the "optimal" k-mer size within a specified range.
  • For example, -k 20-60:5 would find the k-mer size, between 20 and 60 at a step size of 5, that yields the largest number of unique k-mers according to ntCard histograms.
  • Turns on the -ntcard option.

short-read assembly changes

• Initial support for single-end reads.
  • Supply your reads with the -left option and the orientation (ie. -rcl) and strandedness (ie. -ss) accordingly.
• New -mergepool option for merging pooled assemblies.
  • Used in conjunction with the -pool option.
  • Turns off the -norr option.
• Fixed potential memory leak during transcript assembly stage.
• Reduced potential Bloom filter false-positives for paired k-mers.
• Improved overall assembly sensitivity.
• Fixed several bugs.

nanopore reads assembly changes

• Fixed potential memory leak during transcript assembly stage.
• Trim reverse-complement artifacts.

RNA-Bloom v1.2.0

general changes

• significant reduction in overall memory usage for Bloom filters
• new option -savebf to save graph Bloom filters to disk
• new option -debug to output seed sequences in FASTA headers of assembled fragments and transcripts
• supports multi-line records in input FASTA files
• reduced default value for -fpr to 0.01
• increased initial Bloom filter sizes when -ntcard or -nk are not used
• various bug-fixes

short-read assembly changes

• reduced misassemblies
• new option -ref for reference-guided assembly with reference transcriptome(s)
• redundancy reduction in transcript assembly is turn on by default
  • reduces assembly redundancy and overlaps assembled transcripts
  • results are written to *.transcripts.nr.fa
  • minimap2 is required in the PATH
  • new option -norr to turn off redundancy reduction
  • option -nr was removed

nanopore reads assembly changes

• removed dependency on MINIASM
• supports direct cDNA-seq data (see README)
• inter-cluster assembly
  • results are written to *-transcripts.fa

RNA-Bloom v1.1.1

• Fixes compatibility issue with Java 11 (See #1)
• -long option now presets other relevant options (ie. -k, -e, -indel, -p) for long read assembly
• -stage option is supported in long read assemblies (See README for details)
• Significantly improved runtime for long read assembly
• New -mmaln option to generate CIGAR for long read overlaps in minimap2 (Note that this was the default behavior in RNA-Bloom version 1.1.0)