Benchmarking BamScale Across Step1, GAlignments, and SeqQual Workloads

Overview

This article combines the final benchmark runs used for the current BamScale benchmark summary:

step1 and galignments from run_20260320_133141
seqqual from run_20260320_162359

These runs were generated with the same server-first benchmark harness, the same balanced profile family, the same deterministic case order, and the same worker/thread budget policy. Seqqual is reported separately because it includes both the fair compatibility track and the optimized compact track.

Data Loading

Benchmark Provenance

Step1 and GAlignments run: run_20260320_133141
SeqQual run: run_20260320_162359
Step1/GAlignments results directory: /home/runner/work/BamScale/BamScale/inst/benchmarks/run_20260320_133141
SeqQual results directory: /home/runner/work/BamScale/BamScale/inst/benchmarks/run_20260320_162359

Run	Workloads	Profile	CPU	Logical cores	RAM (GB)	Successful cases
run_20260320_133141	step1, galignments	balanced	Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz	96	723.6	32
run_20260320_162359	seqqual	balanced	Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz	96	723.6	26

Methods Rationale

This benchmark suite covers three distinct access patterns:

step1: alignment-metadata extraction without sequence/quality materialization. This is a good proxy for BAM filtering, fragment-length profiling, and fragment-distribution QC.
galignments: construction of alignment objects suitable for Bioconductor workflows centered on GenomicAlignments.
seqqual: full sequence and quality extraction. This is reported in two BamScale modes:
- fair compatibility mode, which is the closest comparator to Rsamtools::scanBam
- optimized compact mode, which reduces internal overhead and should be interpreted as an optimized BamScale track rather than a strict apples-to-apples replacement for the compatibility path

Across all runs, the benchmark design emphasizes:

deterministic case order
balanced BamScale worker/thread plans
explicit comparator baselines
single-file and multi-file reporting
fixed 48-thread budget for multi-file plans

Input Files

The same underlying four selected BAMs were used for both runs, with repeated files allowed to populate the 12-file multi-file benchmark set.

file	source	selected_for_single	selected_for_multi	size_mb	has_index
/home/chiragp/.cache/R/ExperimentHub/134ab745547e62_2073	chipseqDBData	TRUE	TRUE	548.3	TRUE
/home/chiragp/.cache/R/ExperimentHub/134ab7270c5da5_2072	chipseqDBData	FALSE	TRUE	320.4	TRUE
/home/chiragp/.cache/R/ExperimentHub/134ab721bfa655_2071	chipseqDBData	FALSE	TRUE	305.2	TRUE
/home/chiragp/.cache/R/ExperimentHub/134ab7231ef47d_2074	chipseqDBData	FALSE	TRUE	227.3	TRUE

Reference Counts

run	scenario	workload	n_files	n_records	total_mb
run_20260320_133141	multi	galignments	12	77543925	4203.4
run_20260320_133141	multi	step1	12	132482148	4203.4
run_20260320_133141	single	galignments	1	4670364	548.3
run_20260320_133141	single	step1	1	16675372	548.3
run_20260320_162359	multi	seqqual	12	132482148	4203.4
run_20260320_162359	single	seqqual	1	16675372	548.3

Best-Observed Summary

Scenario	Workload	Track	Method family	Method	Plan	Median time (s)
multi	galignments	fair	BamScale	BamScale (balanced budget)	1x48	61.204
multi	galignments	fair	GenomicAlignments	GenomicAlignments::readGAlignments + BiocParallel	12x1	90.777
multi	seqqual	fair	BamScale	BamScale (balanced budget)	1x48	181.277
multi	seqqual	fair	Rsamtools	Rsamtools::scanBam + BiocParallel	12x1	134.667
multi	seqqual	optimized	BamScale	BamScale (compact seqqual budget)	12x4	175.451
multi	step1	fair	BamScale	BamScale (balanced budget)	1x48	68.469
multi	step1	fair	Rsamtools	Rsamtools::scanBam + BiocParallel	12x1	107.882
single	galignments	fair	BamScale	BamScale	1x48	4.047
single	galignments	fair	GenomicAlignments	GenomicAlignments::readGAlignments	1x1	10.568
single	seqqual	fair	BamScale	BamScale	1x48	26.295
single	seqqual	fair	Rsamtools	Rsamtools::scanBam	1x1	24.327
single	seqqual	optimized	BamScale	BamScale (compact seqqual)	1x48	15.026
single	step1	fair	BamScale	BamScale	1x48	8.035
single	step1	fair	Rsamtools	Rsamtools::scanBam	1x1	15.403

BamScale-versus-Comparator Fold Change

Scenario	Workload	Track	Comparator family	Comparator method	Comparator plan	Comparator time (s)	BamScale method	BamScale plan	BamScale time (s)	Comparator/BamScale	BamScale % faster
single	galignments	fair	GenomicAlignments	GenomicAlignments::readGAlignments	1x1	10.568	BamScale	1x48	4.047	2.611	61.705
single	step1	fair	Rsamtools	Rsamtools::scanBam	1x1	15.403	BamScale	1x48	8.035	1.917	47.835
multi	galignments	fair	GenomicAlignments	GenomicAlignments::readGAlignments + BiocParallel	12x1	90.777	BamScale (balanced budget)	1x48	61.204	1.483	32.578
multi	step1	fair	Rsamtools	Rsamtools::scanBam + BiocParallel	12x1	107.882	BamScale (balanced budget)	1x48	68.469	1.576	36.534
single	seqqual	fair	Rsamtools	Rsamtools::scanBam	1x1	24.327	BamScale	1x48	26.295	0.925	-8.088
single	seqqual	optimized	Rsamtools	Rsamtools::scanBam	1x1	24.327	BamScale (compact seqqual)	1x48	15.026	1.619	38.232
multi	seqqual	fair	Rsamtools	Rsamtools::scanBam + BiocParallel	12x1	134.667	BamScale (balanced budget)	1x48	181.277	0.743	-34.611
multi	seqqual	optimized	Rsamtools	Rsamtools::scanBam + BiocParallel	12x1	134.667	BamScale (compact seqqual budget)	12x4	175.451	0.768	-30.285

Single-File `step1`

Method family	Method	Plan	Median time (s)
BamScale	BamScale	1x48	8.0350
BamScale	BamScale	1x24	8.9795
BamScale	BamScale	1x12	9.3775
BamScale	BamScale	1x4	11.0565
Rsamtools	Rsamtools::scanBam	1x1	15.4030
BamScale	BamScale	1x1	16.3500

Single-File `galignments`

Method family	Method	Plan	Median time (s)
BamScale	BamScale	1x48	4.0470
BamScale	BamScale	1x24	4.1920
BamScale	BamScale	1x12	4.7585
BamScale	BamScale	1x4	6.6230
GenomicAlignments	GenomicAlignments::readGAlignments	1x1	10.5680
BamScale	BamScale	1x1	11.0405

Multi-File `step1`

Multi-File `galignments`

Single-File `seqqual`

Method	Plan	Median time (s)
BamScale (compact)	1x48	15.0265
BamScale (compact)	1x24	16.4440
BamScale (compact)	1x12	17.0985
BamScale (compact)	1x4	19.8780
Rsamtools::scanBam	1x1	24.3275
BamScale (compact)	1x1	25.5300
BamScale (compatible)	1x48	26.2950
BamScale (compatible)	1x12	27.6200
BamScale (compatible)	1x24	28.3390
BamScale (compatible)	1x4	31.6780
BamScale (compatible)	1x1	35.6830

Multi-File `seqqual`

Compact-versus-Compatible `seqqual`

Scenario	Workload	Plan	Compatible time (s)	Compact time (s)	Compatible/Compact
multi	seqqual	1x48	181.277	186.321	0.973
multi	seqqual	2x24	573.752	337.596	1.700
multi	seqqual	4x12	404.183	241.524	1.673
multi	seqqual	8x6	361.750	210.290	1.720
multi	seqqual	12x4	270.889	175.451	1.544
single	seqqual	1x1	35.683	25.530	1.398
single	seqqual	1x4	31.678	19.878	1.594
single	seqqual	1x12	27.620	17.098	1.615
single	seqqual	1x24	28.339	16.444	1.723
single	seqqual	1x48	26.295	15.026	1.750

Interpretation

The benchmark shows a consistent pattern across the final runs:

step1 and galignments are the strongest BamScale workloads.
For these workloads, the best BamScale plans are usually low-worker, high-thread configurations, indicating that within-file multithreading contributes more than aggressively increasing the number of file workers.
Seqqual is more difficult because sequence and quality extraction amplifies output-construction overhead.
BamScale compact mode narrows this gap substantially and produces the strongest single-file seqqual result, but the best multi-file seqqual comparator remains faster.

In practical terms, these results support the following guidance:

use BamScale when metadata extraction or alignment-object construction is a bottleneck in BAM-heavy Bioconductor workflows
favor low-worker, high-thread plans when traversing one or a small number of large BAM files
interpret compact seqqual as an optimized throughput path rather than as a strict compatibility benchmark

Session Information

## R version 4.6.0 (2026-04-24)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.4 LTS
## 
## Matrix products: default
## BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0
## 
## locale:
##  [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8       
##  [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8   
##  [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C          
## [10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C   
## 
## time zone: UTC
## tzcode source: system (glibc)
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] scales_1.4.0     ggplot2_4.0.3    tidyr_1.3.2      dplyr_1.2.1     
## [5] readr_2.2.0      BiocStyle_2.40.0
## 
## loaded via a namespace (and not attached):
##  [1] bit_4.6.0           gtable_0.3.6        jsonlite_2.0.0     
##  [4] crayon_1.5.3        compiler_4.6.0      BiocManager_1.30.27
##  [7] tidyselect_1.2.1    parallel_4.6.0      jquerylib_0.1.4    
## [10] systemfonts_1.3.2   textshaping_1.0.5   yaml_2.3.12        
## [13] fastmap_1.2.0       R6_2.6.1            labeling_0.4.3     
## [16] generics_0.1.4      knitr_1.51          tibble_3.3.1       
## [19] bookdown_0.46       desc_1.4.3          RColorBrewer_1.1-3 
## [22] bslib_0.10.0        pillar_1.11.1       tzdb_0.5.0         
## [25] rlang_1.2.0         cachem_1.1.0        xfun_0.57          
## [28] S7_0.2.2            fs_2.1.0            sass_0.4.10        
## [31] bit64_4.8.0         cli_3.6.6           withr_3.0.2        
## [34] pkgdown_2.2.0       magrittr_2.0.5      digest_0.6.39      
## [37] grid_4.6.0          vroom_1.7.1         hms_1.1.4          
## [40] lifecycle_1.0.5     vctrs_0.7.3         evaluate_1.0.5     
## [43] glue_1.8.1          farver_2.1.2        ragg_1.5.2         
## [46] rmarkdown_2.31      purrr_1.2.2         tools_4.6.0        
## [49] pkgconfig_2.0.3     htmltools_0.5.9