automatic model selection, bps not matching with expected

[aistBMRG]

Hi,

Small question regarding the automatic model selection with dorado.

I ran sequencing at 260 bases per second but dorado appears to select 400bps for basecalling.

Fyi, the POD5 (using inspect command) contains the following: 'sample_frequency': '4000', 'selected_speed_bases_per_second': '260'

Now, I am running dorado (0.6.2) as follows:

./bin/dorado basecaller sup ./pod5/

This then returns: "downloading dna_r10.4.1_e8.2_400bps_sup@v4.1.0 with foundation"

Would it be possible to clarify?

Thanks in advance.

Dieter

[HalfPhoton]

Are you able to report the flowcell and kit codes from your input data?

pod5 inspect debug <data.pod5> | grep -E "flow_cell_product_code|sequencing_kit"

# Should look like:
        flow_cell_product_code: <flowcell_product_code>
        sequencing_kit: <sequencing_kit>

Kind regards, Rich

[aistBMRG]

Hi Rich,

Please find below the info:

context_tags: {'barcoding_enabled': '0', 'basecall_config_filename': 'dna_r10.4.1_e8.2_260bps_hac.cfg', 'experiment_duration_set': '4320', 'experiment_type': 'genomic_dna', 'local_basecalling': '1', 'package': 'bream4', 'package_version': '7.4.8', 'sample_frequency': '4000', 'selected_speed_bases_per_second': '260', 'sequencing_kit': 'sqk-lsk114'} flow_cell_product_code: FLO-MIN114 sequencing_kit: sqk-lsk114

Note that this was already base-called in real-time but just testing newer models.

Thanks for your help.

Dieter

[HalfPhoton]

Dorado expects the flowcell product code to contain the _260 suffix for 260bps data.

I'll report this back to understand why this isn't being set - we'll also look into detecting the 'selected_speed_bases_per_second': '260', ' value.

In your case - please manually download and select the appropriate 260bps model as the auto model selection in not working here.

[aistBMRG]

Ok, thanks for the clarification.

Best,

Dieter