PacificBiosciences · Sep 25, 2019 · Feb 18, 2020 · Feb 24, 2020 · Mar 12, 2020 · Mar 12, 2020
diff --git a/.DS_Store b/.DS_Store
diff --git a/LICENSE b/LICENSE
@@ -1,4 +1,4 @@
-Copyright (c) 2018, Pacific Biosciences of California, Inc.
+Copyright (c) 2022, Pacific Biosciences of California, Inc.
 
 All rights reserved.
 

diff --git a/README.md b/README.md
@@ -1,136 +1,35 @@
-<h1 align="center"><img width="300px" src="doc/img/isoseq3.png"/></h1>
-<h1 align="center">IsoSeq v3</h1>
+<h1 align="center"><img width="300px" src="doc/img/isoseq.png"/></h1>
+<h1 align="center">IsoSeq</h1>
 <p align="center">Scalable De Novo Isoform Discovery</p>
 
 ***
 
-*IsoSeq v3* contains the newest tools to identify transcripts in
-PacBio single-molecule sequencing data.
-Starting in SMRT Link v6.0.0, those tools power the
-*IsoSeq GUI-based analysis* application.
-A composable workflow of existing tools and algorithms, combined with
-a new clustering technique, allows to process the ever-increasing yield of PacBio
-machines with similar performance to *IsoSeq* versions 1 and 2.
+*IsoSeq* contains the newest tools to identify transcripts in PacBio
+single-molecule sequencing data. Starting in SMRT Link v6.0.0, those tools power
+the *IsoSeq GUI-based analysis* application. A composable workflow of existing
+tools and algorithms, combined with new clustering techniques, allows to process
+the ever-increasing yield of PacBio machines. Starting with version 3.4, support
+for UMI and cell barcode based deduplication has been added. Version 4.0 adds a
+new `cluster2` tool that enables clustering of hundreds of millions of HiFi
+reads.
+
+## Announcement
+The binary has been renamed from `isoseq3` to `isoseq` to enable major version
+changes. Bioconda will still generate a `isoseq3` softlink. The old bioconda
+`isoseq3` package will automatically install the latest `isoseq` package.
 
 ## Availability
-Latest version can be installed via bioconda package `isoseq3`.
+Latest version can be installed via bioconda package `isoseq`.
 
 Please refer to our [official pbbioconda page](https://github.com/PacificBiosciences/pbbioconda)
 for information on Installation, Support, License, Copyright, and Disclaimer.
 
-## Specific Version Documentation
-
- * [Version 3.2, SMRT Link 8.0](README_v3.2.md)
- * [Version 3.1, SMRT Link 7.0](README_v3.1.md)
- * [Version 3.0, SMRT Link 6.0](README_v3.0.md)
-
-## Changelog
- * **3.2.2**
-   * Fix `polish` not generating fasta/q output. This bug was introduced in v3.2.0
- * 3.2.1
-   * Fix a gff index 1-off bug in `collapse`
-   * We have removed implicit dependencies from the bioconda recipe. Please install `pbccs`, `lima`, and `pbcoretools` as needed.
- * 3.2.0
-   * Add `collapse` step for aligned transcript BAM input
-   * Enable CCS-only workflow `cluster --use-qvs`
-   * Add `refine --min-polya-length`
-   * Add `cluster --singletons` to output unclustered FLNCs; potential sample prep artifacts!
-   * Fix minimap2 bugs. Outputs might change slightly.
- * 3.1.2
-   * Reduce `polish` memory footprint
- * 3.1.1
-   * Edge case fix where `polish` would not finish and stale
-   * Improve `polish` run time for large scale datasets (> 1M CCS)
-   * Improve `polish` result quality
- * 3.1.0
-   * We outsourced the poly(A) tail removal and concatemer detection into a new tool
-called `refine`. Your custom `primers.fasta` is used in this step to detect
-concatemers.
-
-## FAQ
-### Why IsoSeq v3 and not the established versions 1 or 2?
-The ever-increasing throughput of the Sequel system gave rise to the need for a
-scalable software solution that can handle millions of CCS reads, while
-maintaining sensitivity and accuracy. Internal benchmarks have shown that
-*IsoSeq v3* is orders of magnitude faster than currently employed solutions and
-[SQANTI](https://bitbucket.org/ConesaLab/sqanti) attributes *IsoSeq v3* a higher
-number of perfectly annotated isoforms:
-
-<img width="1000px" src="doc/img/isoseq3-performance.png"/>
-
-Additional benefit, single linux binary that requires no dependencies.
-
-### Why is the number of transcripts much lower with IsoSeq3?
-Even though we also observe fewer polished transcripts with *IsoSeq v3*, the
-overall quality is much higher. Most of the low-quality transcripts are lost in the
-demultiplexing step. *Isoseq v1/2 classify* is too relaxed and is not filtering
-junk molecules to a satisfactory level. In fact, *lima* calls are spot on and
-effectively removes most molecules that are wrongly tagged, as in two 5' or two
-3' primers. Only a proper 5' and 3' primer pair allows to identify a full-length
-transcript and its orientation.
-
-
-### I can't find the *classify* step
-Starting with version 3.1, *classify* functionality has been split into two tools.
-Removal of (barcoded) primers is performed with PacBio's standard demultiplexing
-tool *lima*. *Lima* does not remove poly(A) tails, nor detects concatemers.
-For this, `isoseq3 refine` generates FLNC reads.
-
-For version 3.0, poly(A) tail removal and concatemer detection is performed in
-`isoseq3 cluster`
-
-### My sample has poly(A) tails, how can I remove them?
-Use `--require-polya` for `isoseq3 refine`.
-This filters for FL reads that have a poly(A) tail
-with at least 20 base pairs and removes identified tail.
-
-### How long will it take until my data has been processed?
-There is no ETA feature. Depending on the sample type, whole transcriptome
-or targeted amplification, run time varies. The same number of reads from a
-whole transcriptome sample can finish clustering in minutes, whereas a single
-gene amplification of 10kb transcripts can take a couple of hours.
-
-### Which clustering algorithm is used?
-In contrast to its predecessors, *IsoSeq v3* does not rely on NP-hard clique
-finding, but uses a hierarchical alignment strategy with `O(N*log(N))`.
-Recent advances in rapid alignment of long reads make this this approach
-feasible.
-
-### How many CCS reads are used for the unpolished cluster sequence representation?
-*Cluster* uses up to 10 CCS reads to generate the unpolished cluster consensus.
-
-### How many subreads are used for polishing?
-*Polish* uses up to 60 subreads to polish the cluster consensus.
-
-### When are two reads clustered?
-*IsoSeq v3* deems two reads to stem from the same transcript, if they meet
-following criteria:
-
-<img width="1000px" src="doc/img/isoseq3-similar-transcripts.png"/>
-
-There is no upper limit on the number of gaps.
-
-### BAM tags explained
-Following BAM tags are being used:
-
- - `ib` Barcode summary: triplets delimited by semicolons, each triplet contains two barcode indices and the ZMW counts, delimited by comma. Example: `0,1,20;0,3,5`
- - `im` ZMW names associated with this isoform
- - `is` Number of ZMWs associated with this isoform
- - `iz` Maximum number of subreads used for polishing
- - `rq` Predicted accuracy for polished isoform
-
- Quality values are capped at `93`.
-
-### What SMRTbell designs are possible?
-
-PacBio supports three different SMRTbell designs for IsoSeq libraries.
-In all designs, transcripts are labelled with asymmetric primers,
-whereas a poly(A) tail is optional. Barcodes may be optionally added.
+## Latest Version
+Version **4.3.0**: [Full changelog here](docs/changelog)
 
-<img width="600px" src="doc/img/isoseq3-barcoding.png"/>
+## Workflow Documentation
 
-### The binary does not work on my linux system!
-Binaries require **SSE4.1 CPU support**; CPUs after 2008 (Penryn) include it.
+ * Visit [isoseq.how](https://isoseq.how) for the latest documentation
 
 ## DISCLAIMER
 

diff --git a/README_v3.0.md b/README_v3.0.md