datastore might fit your case.
If your PC has enought memory (RAM) to store all tsv files in workspace at once, you can use readall function.
For example, suppose tsv files are located in data directory and each tsv has column name col1, you can calculate the mean by the following.
ds = datastore('data/*.tsv', 'Type', 'tabulartext', 'FileExtensions', '.tsv');
t = readall(ds);
m = mean(t.col1);
Or, if the RAM is not sufficient, you can read by a chunk.
ds = datastore('data/*.tsv', 'Type', 'tabulartext', 'FileExtensions', '.tsv');
sums = [];
counts = [];
while hasdata(ds)
t = read(ds);
sums(end+1) = sum(t.col1);
counts(end+1) = length(t.col1);
end
m = sum(sums)/sum(counts);
alternatively, you can also use tall array.
ds = datastore('data/*.tsv', 'Type', 'tabulartext', 'FileExtensions', '.tsv');
tt = tall(ds);
m = gather(mean(tt.col1));
