6 Performance

The original workflow benchmark used 12 threads across RNA-seq and Ribo-seq datasets from three species.

Species	Dataset	Library	Sample number	Sample size	Index building elapsed time	Index disk usage	Preprocessing & alignment elapsed time	Preprocessing disk usage	RiboParser elapsed time	RiboParser disk usage
S. cerevisiae	GSE67387	Ribo-seq	6	32 G	38 s	357 M	43 m 21 s	30 G	1 h 26 m 23 s	3.6 G
S. cerevisiae	GSE67387	RNA-seq	6	17 G	38 s	357 M	32 m 52 s	26 G	37 m 42 s	2.8 G
M. musculus	GSE114064	Ribo-seq	6	43 G	59 m 8 s	36 G	50 m 27 s	31 G	32 m 3 s	7.9 G
M. musculus	GSE114064	RNA-seq	6	60 G	59 m 8 s	36 G	4 h 14 m 45 s	62 G	29 m 30 s	7.6 G
H. sapiens	GSE131650	Ribo-seq	6	42 G	1 h 55 m 56 s	44 G	2 h 11 m 42 s	29 G	2 h 18 m 57 s	14 G
H. sapiens	GSE131650	RNA-seq	6	54 G	1 h 55 m 56 s	44 G	1 h 15 m 15 s	30 G	40 m 35 s	11 G

Recommendation

Use Linux-based systems for production analysis. For large projects, run sample-level steps using SLURM array jobs and merge results only after all samples complete.