diff --git a/README.md b/README.md
index f67bfed..ecaac82 100644
--- a/README.md
+++ b/README.md
@@ -10,3 +10,5 @@ panct is a collection of tools for working pangenomes
 Homepage: [https://panct.readthedocs.io/](https://panct.readthedocs.io/)
 
 Visit our homepage for installation and usage instructions.
+
+Precomputed region-level complexity scores (at 50kb, 100kb, and 1Mb) resolution for hg38 are available in [precomputed-scores](./precomputed-scores).
\ No newline at end of file
diff --git a/precomputed-scores/README.md b/precomputed-scores/README.md
new file mode 100644
index 0000000..2a9ae3a
--- /dev/null
+++ b/precomputed-scores/README.md
@@ -0,0 +1,44 @@
+# Precomputed pangenome complexity scores
+
+Scores are computed for 50kb, 100kb, or 1Mb windows on hg38 in the files:
+
+* `hprc-v1.1-mc-grch38_complexity_50000.tab`
+* `hprc-v1.1-mc-grch38_complexity_100000.tab`
+* `hprc-v1.1-mc-grch38_complexity_1000000.tab`
+
+Columns give:
+
+* chrom, start, end: genomic location of the window in hg38
+* numnodes: number of minigraph-cactus nodes identified in the window
+* total_length: sum of the lengths of all nodes in the window (note this will usually be close to but not exactly equal to the length of the hg38 window, since it includes the lengths of non-reference nodes)
+* numwalks: number of walks identified through the subgraph. This is based on output of the `query` command from gbz-base and is in some cases more than the total number of assemblies used to build the graph.
+* sequniq-normwalk:  sum_n  len(n)*p_n*(1-p_n)/L where L is the average walk length
+* sequniq-normnode: sum_n  len(n)*p_n*(1-p_n)/L where L is the average node length
+
+Scores of None indicate no walks were identified through the region.
+
+## Computing scores
+
+Scores were generated using the following commands:
+
+1. Make windows of different sizes across hg38
+```
+for window in 50000 100000 1000000
+do
+	bedtools makewindows -g hg38.txt -w ${window} > windows/hg38_windows_${window}.bed
+done
+```
+
+2. Compute complexity scores for each window based on the HPRC minigraph-cactus v1 graph
+
+Scores are based on hprc-v1.1-mc-grch38 available here: https://s3-us-west-2.amazonaws.com/human-pangenomics/pangenomes/freeze/freeze1/minigraph-cactus/hprc-v1.1-mc-grch38/hprc-v1.1-mc-grch38.gbz
+```
+for window in 50000 100000 1000000
+do
+	panct complexity \
+		--region windows/hg38_windows_${window}.bed \
+		--out hprc-v1.1-mc-grch38_complexity_${window}.tab \
+		--metrics sequniq-normwalk,sequniq-normnode \
+		../testdata/hprc-v1.1-mc-grch38.gbz
+done
+```
\ No newline at end of file
diff --git a/precomputed-scores/hg38.txt b/precomputed-scores/hg38.txt
new file mode 100644
index 0000000..4509a88
--- /dev/null
+++ b/precomputed-scores/hg38.txt
@@ -0,0 +1,25 @@
+chr1	248956422
+chr2	242193529
+chr3	198295559
+chr4	190214555
+chr5	181538259
+chr6	170805979
+chr7	159345973
+chrX	156040895
+chr8	145138636
+chr9	138394717
+chr11	135086622
+chr10	133797422
+chr12	133275309
+chr13	114364328
+chr14	107043718
+chr15	101991189
+chr16	90338345
+chr17	83257441
+chr18	80373285
+chr20	64444167
+chr19	58617616
+chrY	57227415
+chr22	50818468
+chr21	46709983
+chrM	16569
\ No newline at end of file
diff --git a/precomputed-scores/hprc-v1.1-mc-grch38_complexity_100000.tab.gz b/precomputed-scores/hprc-v1.1-mc-grch38_complexity_100000.tab.gz
new file mode 100644
index 0000000..9ce559a
Binary files /dev/null and b/precomputed-scores/hprc-v1.1-mc-grch38_complexity_100000.tab.gz differ
diff --git a/precomputed-scores/hprc-v1.1-mc-grch38_complexity_1000000.tab.gz b/precomputed-scores/hprc-v1.1-mc-grch38_complexity_1000000.tab.gz
new file mode 100644
index 0000000..f090580
Binary files /dev/null and b/precomputed-scores/hprc-v1.1-mc-grch38_complexity_1000000.tab.gz differ
diff --git a/precomputed-scores/hprc-v1.1-mc-grch38_complexity_50000.tab.gz b/precomputed-scores/hprc-v1.1-mc-grch38_complexity_50000.tab.gz
new file mode 100644
index 0000000..071110d
Binary files /dev/null and b/precomputed-scores/hprc-v1.1-mc-grch38_complexity_50000.tab.gz differ