Bug: nma.pdbs() got error when input pdbs have blank chain id

[Xinqiu Yao]

Following pdbs were obtained from MD setup, which all contain blank chain id

f <- c("gdp_ca.pdb", "gtp_ca.pdb", "open_ca.pdb")
pdbs <- pdbaln(f)
gdp <- read.pdb(f[1])

# nma() is okay
modes <- nma(gdp)
# Building Hessian...       Done in 0.308 seconds.
# Diagonalizing Hessian...  Done in 1.566 seconds.

#nma.pdbs() errors
modes <- nma.pdbs(pdbs)

# Building Hessian...       Done in 0.358 seconds.
# Diagonalizing Hessian...Error in eigen(H, symmetric = TRUE) : infinite or missing #values in 'x'
#In addition: Warning message:
#In read.pdb(fname) : NAs introduced by coercion

Possible reason is read.pdb() and write.pdb(), in which blank chain id are recognized as NA

[Lars Skjærven]

thanks for noting this. it has been fixed in the recent update.

However, the way I make a PDB object in nma.pdbs might not be best. i.e. I use write.pdb and read.pdb. might be a better solution for making a PDB object, but it was easy (and secure)

[Xinqiu Yao]

It seems resolved.