nf-core/proteinfold
Edit

Protein 3D structure prediction pipeline

alphafold2protein-fold-predictionprotein-foldingprotein-sequencesprotein-structure

This is the development version of the pipeline.

Launch development version https://github.com/nf-core/proteinfold

Introduction

This document describes the output produced by the pipeline.

Most of the plots are taken from the MultiQC report, which summarises results at the end of the pipeline.

Pipeline overview

The pipeline is built using Nextflow and predicts protein structures using the following methods:

AlphaFold2
ColabFold - MMseqs2 (API server or local search) followed by ColabFold
ESMFold

See main README.md for a condensed overview of the steps in the pipeline, and the bioinformatics tools used at each step.

The directories listed below will be created in the output directory after the pipeline has finished. All paths are relative to the top-level results directory.

AlphaFold2

Output files

alphafold2/standard/ or alphafold2/split_msa_prediction/ based on the selected mode. It contains the computed MSAs, unrelaxed structures, relaxed structures, ranked structures, raw model outputs, prediction metadata, and section timings. Specifically, <SEQUENCE NAME>_plddt_mqc.tsv presents the pLDDT scores per residue for each of the 5 predicted models.
- top_ranked_structures/<SEQUENCE NAME>.pdb that is the structure with the highest pLDDT score per input (ranked first)
DBs/ that contains symbolic links to the downloaded database and parameter files

Below you can find an indicative example of the TSV file with the pLDDT scores per residue for each of the 5 predicted models produced by AlphaFold2, which is included in the MultiQC report:

Positions	rank_0	rank_1	rank_2	rank_3	rank_4
1	66.17	60.61	60.32	64.20	65.31
2	78.01	74.20	73.11	77.36	78.46
3	82.16	78.16	76.70	80.20	80.68
4	86.03	82.78	81.88	82.19	83.93
5	88.08	84.38	84.73	85.58	87.70
6	89.37	86.06	86.31	86.84	88.52
7	91.27	88.27	88.09	87.01	88.67
8	91.28	89.42	90.17	87.47	90.07
9	93.10	90.09	92.86	90.70	93.41
10	93.23	91.42	93.07	90.13	92.91
11	94.12	92.44	93.00	89.90	92.97
12	95.15	93.63	94.25	92.66	94.38
13	95.09	93.75	94.36	92.54	94.95
14	94.08	92.72	93.43	90.31	93.63
15	94.34	93.77	93.31	91.72	93.57
16	95.56	94.62	94.46	93.55	95.20
17	95.54	94.75	94.65	93.61	95.37
18	93.91	93.89	93.30	91.33	92.95
19	95.48	95.78	94.48	93.95	95.05
20	95.96	95.46	95.14	94.01	95.83
21	94.06	94.06	93.13	91.69	93.54
22	92.98	93.28	91.14	88.80	91.25
23	95.28	95.13	93.39	91.48	93.56
24	93.41	93.38	92.32	89.85	92.40
25	90.88	91.40	88.60	85.67	87.65
26	89.30	88.90	84.58	83.11	84.52
27	91.96	90.95	89.04	86.42	87.77
28	91.20	90.68	88.71	86.43	87.62
29	88.01	87.53	85.83	83.11	84.95
30	81.29	83.72	77.75	75.76	74.84
31	87.14	86.92	82.10	82.32	78.74
32	92.34	90.13	89.04	88.31	86.49
33	91.70	88.94	85.52	85.94	81.75
34	90.11	88.23	84.33	85.47	80.01
35	93.35	91.49	90.60	89.40	87.10
36	94.15	92.47	90.17	90.48	86.77
37	93.40	92.01	86.38	87.84	80.11
38	92.79	89.97	89.31	88.55	85.15
39	94.66	91.29	92.74	90.67	90.30
40	95.98	93.58	94.30	91.69	90.73
41	94.94	92.57	88.31	88.40	80.33
42	92.89	91.03	84.03	85.31	74.66
43	94.54	93.44	86.50	84.91	76.68
44	96.93	95.23	92.42	91.98	86.11
45	94.40	92.27	87.40	89.02	79.44
46	91.74	90.94	81.35	84.88	74.93
47	96.19	94.46	90.51	89.82	84.51
48	94.84	93.04	91.02	91.57	87.72
49	91.23	89.34	86.10	87.63	82.12
50	91.64	89.58	84.93	85.88	79.38

ColabFold

Output files

colabfold/webserver/ or colabfold/local/ based on the selected mode. It contains the computed MSAs, unrelaxed structures, relaxed structures, ranked structures, raw model outputs, prediction metadata, and section timings. Specifically, <SEQUENCE NAME>_plddt_mqc.tsv presents the pLDDT scores per residue for each of the 5 predicted models.
- top_ranked_structures/<SEQUENCE NAME>.pdb that is the structure with the highest pLDDT score per input (ranked first)
DBs/ that contains symbolic links to the downloaded database and parameter files

Below you can find some indicative examples of the output images produced by ColabFold, which are included in the MultiQC report:

Sequence coverage

Alt text

predicted Local Distance Difference Test (pLDDT)

Alt text

Predicted Aligned Error (PAE)

Alt text

ESMFold

Output files

esmfold/default contains the predicted structures. Specifically, <SEQUENCE NAME>_plddt_mqc.tsv presents the pLDDT scores per residue for each of the predicted models.
- top_ranked_structures/<SEQUENCE NAME>.pdb that is the structure with the highest pLDDT score per input (ranked first)
DBs/ that contains symbolic links to the downloaded database and parameter files

Below you can find an indicative example of the TSV file with the pLDDT scores per atom for predicted model produced by ESMFold, which is included in the MultiQC report:

Atom_serial_number	Atom_name	Residue_name	Residue_sequence_number	pLDDT
1	N	VAL	1	44.77
2	CA	VAL	1	47.23
3	C	VAL	1	46.66
4	CB	VAL	1	41.88
5	O	VAL	1	45.75
6	CG1	VAL	1	39.15
7	CG2	VAL	1	39.59
8	N	THR	2	49.89
9	CA	THR	2	51.41
10	C	THR	2	50.21
11	CB	THR	2	43.84
12	O	THR	2	47.36
13	CG2	THR	2	35.32
14	OG1	THR	2	40.12
15	N	VAL	3	51.40
16	CA	VAL	3	54.38
17	C	VAL	3	52.10
18	CB	VAL	3	48.50
19	O	VAL	3	52.58
20	CG1	VAL	3	38.75
21	CG2	VAL	3	39.26
22	N	ASP	4	52.00
23	CA	ASP	4	53.92
24	C	ASP	4	52.33
25	CB	ASP	4	46.82
26	O	ASP	4	51.28
27	CG	ASP	4	42.89
28	OD1	ASP	4	45.89
29	OD2	ASP	4	53.61
30	N	ASP	5	56.10
31	CA	ASP	5	56.97
32	C	ASP	5	55.75
33	CB	ASP	5	50.34
34	O	ASP	5	54.18
35	CG	ASP	5	45.82
36	OD1	ASP	5	50.03
37	OD2	ASP	5	58.01
38	N	LEU	6	56.50
39	CA	LEU	6	58.34
40	C	LEU	6	55.81
41	CB	LEU	6	52.46
42	O	LEU	6	54.42
43	CG	LEU	6	49.17
44	CD1	LEU	6	44.31
45	CD2	LEU	6	47.07
46	N	VAL	7	57.23
47	CA	VAL	7	57.68
48	C	VAL	7	57.39
49	CB	VAL	7	52.74
50	O	VAL	7	56.46

MultiQC report

Output files

multiqc
- <MODE>_multiqc_report.html: A standalone HTML file that can be viewed in your web browser.
- <MODE>_multiqc_data/: Directory containing parsed statistics from the different tools used in the pipeline.
- <MODE>_multiqc_plots/: Directory containing static images from the report in various formats.

MultiQC is a visualisation tool that generates a single HTML report summarising all samples in your project. Most of the pipeline QC results are visualised in the report and further statistics are available within the report data directory.

Results generated by MultiQC collate pipeline QC from AlphaFold2 or ColabFold.

The pipeline has special steps which also allow the software versions to be reported in the MultiQC output for future traceability. For more information about how to use MultiQC reports, see http://multiqc.info.

Pipeline information

Output files

pipeline_info/
- Reports generated by Nextflow: execution_report.html, execution_timeline.html, execution_trace.txt and pipeline_dag.dot/pipeline_dag.svg.
- Reports generated by the pipeline: pipeline_report.html, pipeline_report.txt and software_versions.yml. The pipeline_report* files will only be present if the --email / --email_on_fail parameter’s are used when running the pipeline.
- Reformatted samplesheet files used as input to the pipeline: samplesheet.valid.csv.
- Parameters used by the pipeline run: params.json.

Nextflow provides excellent functionality for generating various reports relevant to the running and execution of the pipeline. This will allow you to troubleshoot errors with the running of the pipeline, and also provide you with other information such as launch commands, run times and resource usage.

nf-core/proteinfold Edit

Introduction

Pipeline overview

AlphaFold2

ColabFold

Sequence coverage

predicted Local Distance Difference Test (pLDDT)

Predicted Aligned Error (PAE)

ESMFold

MultiQC report

Pipeline information

nf-core/proteinfold
Edit