Skip to content

Commit

Permalink
Feature/pass reads (#361)
Browse files Browse the repository at this point in the history 
* Attempt at using pass_reads

* Add message about generating the aligner description

* FIX bug where parameter reads was being overwritten by empty list reads

* DRAFT Feature/minknow6 compatibility (#369)

* Made a proper SVG thought i'd just include it here

* Remove setting the read_number on the result
Add in the correct strand classifications for minknow 6

* Remove Read number from Result and all references to read number

* Update target versions for MinKNOW to 6.0.0 - 6.0.0
Explicitly exit if running on MInKNOW 5.X.X

* Update vendored read until API to v3.5.2

* Update test for conencted base caller version to work

* Pin ont-pybasecall-client-lib at 7.4.12

* Add downgrade to 2024.2.0 to error message

* Fix bug where unblocked_read_ids wasn't being written out

* Chunk tracker uses read_id not read_number, update trackers internal variable names

* bump version to 2024.3.0

* Fix tests for the new MinKNOW version compatibility
Fix tests for the _statistics too not use `read_number`

* Bump summarise version to 0.2.7

* Drop incorrect gotcha

* Feature/deprecate guppy (#374)

* This commit addresses the removal of guppy.py, partial removal of guppy tests, edits to documentation to largely replace guppy with dorado, removes the ont-pyguppy-client-lib and adds a warning to the readme. It also adds some extra classes to the read_until init in targets.

* Remove builtin module for guppy

* Update FAQ questions to focus dorado as well as Guppy

* Change validate caller warning to mention dorado not guppy

* Change readme warnings to Focus dorado and minknow breaking changes more
Add link/requirement for dorado_basecall_server not guppy_basecall_server

* Rename tests to dorado validation tests

* Add dorado to mappy-test install target

* Fix black formatting of accepted chunks list

* rename read_id to read_number on _generate_action
because I may as well

* return action from override as for some pythons it doesn't mutate in place

* add mappy-rs because I keep typing it and not mappy_rs

* Don't clip RF from front, used cached read_id for the result
  • Loading branch information
@Adoni5
Adoni5 authored Aug 26, 2024
1 parent 8e5210e commit 8c013a4
Show file tree
Hide file tree
Showing 46 changed files with 189,104 additions and 554 deletions.
41 changes: 26 additions & 15 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,11 @@ wrong, let us know so we can add you to the GOTCHA hall of fame!
> We also have more detailed documentation for your perusal at https://looselab.github.io/readfish
> [!NOTE]
> Now also see our cool [FAQ](docs/FAQ.md).
> Now also see our cool [FAQ](docs/FAQ.md).
> [!WARNING]
Breaking for any version of `MinKNOW <= 6.0.0`
As of `readfish >=2024.3.0` we no longer support guppy.

readfish is a Python package that integrates with the
[Read Until API](https://github.com/nanoporetech/read_until_api).
Expand All @@ -23,13 +27,13 @@ way most fit for purpose, and a return call can be made to the server to unblock
the read in progress and so direct sequencing capacity towards reads of interest.


**This implementation of readfish requires Guppy version >= 6.0.0 and MinKNOW version core >= 5.0.0 . It will not work on earlier versions.**
**This implementation of readfish requires Dorado server version >= 7.3.9 and MinKNOW version core >= 6.0.0 . It will not work on earlier versions.**

**Since MinKNOW version core >=5.9.0 and Dorado server version >=7.3.9, Dorado requires an alternate library, `ont-pybasecall-client-lib`. We have introduced a new`dorado` module to handle this.**
**To run with earlier versions of MinKNOW please use an earlier version of readfish.**


The code here has been tested with Guppy in GPU mode using GridION Mk1 and
NVIDIA RTX2080 on live sequencing runs and an NVIDIA GTX1080 using playback
The code here has been tested with Dorado in GPU mode using GridION Mk1 and
NVIDIA RTX4090s on live sequencing runs and on MacOSX M2Max using playback
on a simulated run (see below for how to test this).
This code is run at your own risk as it DOES affect sequencing output. You
are **strongly** advised to test your setup prior to running (see below for
Expand All @@ -41,6 +45,7 @@ The following platforms are supported:

- **PromethION** Big Boy
- **P2Solo** Smol Big Boy
- **P2i** Not so Smol Big Boy
- **GridION** Box
- **MinION** Smol Boy

Expand All @@ -50,7 +55,7 @@ The following platforms are supported:
The following OSs are supported:

- **Linux** yay
- **MacOS** boo (Apple Silicon, Only with Dorado)
- **MacOS** boo (Apple Silicon Only)


> [!NOTE]
Expand Down Expand Up @@ -128,7 +133,7 @@ conda activate readfish_dev
| MinKNOW is transitioning from Guppy to Dorado. Until MinKNOW version 5.9 both Guppy and Dorado used ont-pyguppy-client-lib.<br/>As of MinKNOW version 5.9 and Dorado server version 7.3.9 and greater Dorado requires an alternate library, `ont-pybasecall-client-lib`.<br/>The listed `ont-pyguppy-client-lib` or `ont-pybasecaller-client-lib` version may not match the version installed on your system. To fix this, Please see this [issue](https://github.com/LooseLab/readfish/issues/221#issuecomment-1381529409), using the appropriate library. |


[ONT's Guppy GPU](https://community.nanoporetech.com/downloads) should be installed and running as a server.
[ONT's Dorado Basecall Server GPU](https://community.nanoporetech.com/downloads) should be installed and running as a server.

<details style="margin-top: 10px">
<summary><span id="py-ve">Alternatively, install readfish into a python virtual-environment</span></summary>
Expand All @@ -142,8 +147,8 @@ pip install --upgrade pip
# Install our readfish Software
pip install readfish[all]
# Install ont_pyguppy_client_lib that matches your guppy server version. E.G.
pip install ont_pyguppy_client_lib==6.3.8
# Install ont_pybasecall_client that matches your dorado basecall server version. E.G.
pip install ont_pybasecall_client_lib==7.1.2
```

</details>
Expand Down Expand Up @@ -298,8 +303,9 @@ Steps:
```
Change the text between the quotes to point to your downloaded bulk FAST5 file.
<!-- end-obsolete -->
1. Optional, If running GUPPY in GPU mode, set the parameter `break_reads_after_seconds = 1.0`
1. Optional, If running Dorado in GPU mode, you can set the parameter `break_reads_after_seconds = 1.0`
to `break_reads_after_seconds = 0.4`. This results in a smaller read chunk. For R10.4 this is not required but can be tried. For adaptive sampling on PromethION, this should be left at 1 second.
1. In MinKNOW >= 6.0.0 this value defaults to 0.8 which is a reasonable balance.
1. In the MinKNOW GUI, right click on a sequencing position and select `Reload Scripts`.
Your version of MinKNOW will now playback the bulkfile rather than live sequencing.
1. Start a sequencing run as you would normally, selecting the corresponding flow
Expand Down Expand Up @@ -335,13 +341,10 @@ Note: The plots here are generated from running readfish unblock-all on an Apple
<details style="margin-top: 10px">
<summary id="testing-basecalling-and-mapping"><h3 style="display: inline;">Testing base-calling and mapping</h3></summary>

To test selective sequencing you must have access to either a
[guppy basecall server](https://community.nanoporetech.com/downloads/guppy/release_notes) (>=6.0.0) or a [dorado basecall server](https://community.nanoporetech.com/downloads/dorado/release_notes).
To test selective sequencing you must have access to a [dorado basecall server](https://community.nanoporetech.com/downloads/dorado/release_notes).

and a readfish TOML configuration file.

NOTE: guppy and dorado are used here interchangeably as the basecall server. Dorado is gradually replacing guppy. All readfish code is compatible with Guppy >=6.0.0 and dorado >=0.4.0

1. First make a local copy of the example TOML file:
```console
curl -O https://raw.githubusercontent.com/LooseLab/readfish/master/docs/_static/example_tomls/human_chr_selection.toml
Expand Down Expand Up @@ -532,7 +535,7 @@ These may or may not (!) be mistakes we have made already...
- Double check your reference file is in the correct location.
- Double check your targets exist in that reference file.
- Double check your targets are correctly formatted with contig name matching the record names in your reference (Exclude description - i.e the contig name up to the first whitespace).
1. **Where has my reference gone?** If you are using a _live TOML file - e.g running iter_align or iter_cent, the previous reference MMI file is deleted when a new one is added. This obviously saves on disk space use(!) but can lead to unfortunate side effects - i.e you delete your MMI file. These can of course be recreated but user **beware**.


Happy readfish-ing!

Expand All @@ -559,6 +562,14 @@ And for our Awesome Logo please checkout out [@tim_bassford](https://twitter.com

<!-- start-changelog -->
# Changelog
## 2024.3.0
### This release is breaking for ALL versions of `MinKNOW <= 6` and no longer supports Guppy.

1. Introducing support for MinKNOW >=6.0.0 and deprecating support for earlier versions.
1. Removing support for legacy guppy base caller and only supporting Dorado in future.
1. Optimising batch sending to the base caller through the use of `pass_reads` rather than `pass_read`
1. Adding the new strand classifications as used by MinKNOW, including strand2 and short.

## 2024.2.0
1. Add a dorado base-caller which addressed issue [#347](https://github.com/LooseLab/readfish/issues/347) - chiefly in Dorado 7.3.9 ONT have moved to `ont-pybasecall-client-lib`,
and connections from `ont_pyguppy_client_lib` raise `Connection error. ... LOAD_CONFIG. Reply: INVALID_PROTOCOL` [(#344)](https://github.com/LooseLab/readfish/pull/344)
Expand Down
Loading

0 comments on commit 8c013a4

Please sign in to comment.